The Steel Construction Institute
Steel Bridge Group:
Guidance Notes on Best P
in Steel Bridge Con
The Steel Construction Institute develops and promotes the effective use of steel in
construction. It is an independent, membership based organisation.
SCl's research and development activities cover many aspects of steel construction including
multi-storey construction, industrial buildings, light steel framing systems and modular
construction, development of design guidance on the use of stainless steel, fire engineering,
bridge and civil engineering, offshore engineering, environmental studies, value engineering, and
development of structural analysis systems and information technology.
Membership is open t o all organisations and individuals that are concerned with the use of steel
in construction. Members include designers, contractors, suppliers, fabricators, academics and
government departments in the United Kingdom, elsewhere in Europe and in countries around
the world. The SCI is financed b y subscriptions from its members, revenue from research
contracts and consultancy services, publication sales and course fees.
The benefits of corporate membership include access t o an independent specialist advisory
service and free issue of SCI publications as soon as they are produced. A Membership
Information Pack is available on request from the Membership Manager.
The Steel Construction Institute, Silwood Park, Ascot, Berkshire, SL5 7QN.
Telephone: (01 344) 623345, Fax: (01 344) 622944.
World Wide Web site: http://www.steel-sci.org
,
The British Constructional Steelwork Association Limited (BCSA) is the national
organisation for the Constructional Steelwork Industry: its Member companies
undertake the design, fabrication and erection of steelwork for all forms of
construction in building and civil engineering. Associate Members are those
principal companies involved in the purchase, design or supply of components,
materials, services, etc. related to the industry.
The principal objectives of the Association are to promote the use of structural steelwork; t o
assist specifiers and clients; t o ensure that the capabilities and activities of the industry are
widely understood and t o provide members with professional services in technical, commercial,
contractual and quality assurance matters.
The British Constructional Steelwork Association Limited, 4 Whitehall Court, Westminster,
London SW1 A 2ES. Telephone: (020) 7 8 3 9 8566, Fax: (020) 7 9 7 6 1634.
World Wide Web site: http://www.bcsa.org
Corus is the largest steel company in Europe and is also a significant
player in the aluminium industry. Corus's multi-metals approach is one of
the unique features of the company. The Corus Construction Centre
brings together an integrated team of experts t o provide a one-stop-shop
for advice on the use of steel and aluminium in construction. The multidisciplinary team includes civil engineers, structural engineers, architects and construction
economists. The Corus Construction Centre can provide advice on the full range of Corus
products and provide help to specifiers with any queries related t o structural steelwork design.
It also acts as a point of contact with the sales function and Corus Technology. A series of
publications is available dealing with the steel products and their use.
Corus Construction Centre, PO Box 1, Brigg Road, Scunthorpe, North Lincolnshire, D N I 6 1BP.
Telephone: (01724) 405060 Fax: (01724) 404224.
World Wide Web site: http://www.corusconstruction.com
ADDITIONAL NOTES
Guidance Notes on Best Practice in Steel Bridge Construction (P185)
This pack contains 19 new Guidance Notes and 3 revised notes (GN 1.01, GN 2.04 and
GN 9.01). In addition, it also includes, amended inner cover page (full colour),
preliminary pages (i-viii) and 9 contents pages for the individual Sections.
Please substitute the revised pages and the amended pages and add the new notes in the
relevant Sections in your folder.
Further notes will be issued in due course, including the 4 notes that are ‘missing’ from
the numbering sequence. When these become available, we would like to contact you.
Please fill in your details in the form provided below and post or fax it to the address
given below:
Dr F Awan
The Steel Construction Institute
Silwood Park
Ascot
Berkshire
SL5 7QN
Tel: 01344 623345
Fax: 01344 628211
[Retain this half for your reference]
x ......................................................................................
Steel Bridge Group:
Guidance Notes on Best Practice in Steel Bridge Construction (P185)
REGISTRATlON FORM
To: Dr F Awan, Publications Manager, The Steel Construction Institute.
Please include my name and address in the database set up to provide further information
to those who have received the SCI publication P185.
(This database will not be used for any other purpose).
Name
Position
Company
Address
Town/City
Postcode
Telephone
Fax
Date when publication received
Membership No. (if applicable)
P185 Issue 2 Notes
SCI PUBLICATION P1 85
Steel Bridge Group:
Guidance Notes on Best Practice
in Steel Bridge Construction
(Second issue, September 2000)
Edited by:
J E EVANS BSc(Eng1, ACGI, DIC, CEng, FICE, FWeldl
D
c ILES MSc, ACGI, DIC, CEng, MICE
Steel Bridge Group is a steel industry sector committee
supported by BCSA, Corus and SCI
Published by:
The Steel Construction Institute
Silwood Park
Ascot
Berkshire SL5 7QN
Tel: 01344 623345
Fax: 01344 622944
© 1998, 2000 The Steel Construction Institute
Apart from any fair dealing for the purposes of research or private study or criticism or review, as
permitted under the Copyright Designs and Patents Act, 1988, this publication may not be
reproduced, stored or transmitted, in any form or by any means, without the prior permission in
writing of the publishers, or in the case of reprographic reproduction only in accordance with the
terms of the licences issued by the UK Copyright Licensing Agency, or in accordance with the terms
of licences issued by the appropriate Reproduction Rights Organisation outside the UK.
Enquiries concerning reproduction outside the terms stated here should be sent t o the publishers, The
Steel Construction Institute, at the address given on the title page.
Although care has been taken t o ensure, t o the best of our knowledge, that all data and information
contained herein are accurate t o the extent that they relate t o either matters of fact or accepted
practice or matters of opinion at the time of publication, The Steel Construction Institute, the authors
and the reviewers assume no responsibility for any errors in or misinterpretations of such data and/or
information or any loss or damage arising from or related t o their use.
Publications supplied to the Members of the Institute at a discount are not for resale by them.
Publication Number: SCI P185
ISBN 1 85942 061 3
British Library Cataloguing-in-Publication Data.
A catalogue record for this book is available from the British Library.
FOREWORD
This publication presents a collection of separate Guidance Notes on a range of topics
concerning the design and construction of structural steelwork for bridges. I t complements
the publication Specification of structural steelwork for bridges: a model Appendix 18/1,
which was written for use in project specifications and which makes reference to some of
the Guidance Notes in t h i s publication.
Each Note i s based on an initial draft by a member of the Steel Bridge Group (SBG), and
subsequently developed with assistance and comment from other SBG members. The Notes
have been edited by Mr D C Iles (The Steel Construction Institute) and by Mr J E Evans
(Flint and N e i l l Partnership), the chairman of the SBG. In the f i r s t issue (1998), 31 Notes
were published. This second issue includes a further 18 Notes, and three of the earlier
Notes have been revised.
The Steel Bridge Group i s a technical forum that has been established to consider matters
of high-priority interest to the steel bridge construction industry and to suggest strategies
for improving the use of steel in bridgework. At the time of drafting the second issue, the
Group included the following members:
Mr J E Evans (Chairman)
Mr G W Booth
Mr S Chakrabarti
Mr R E Craig
Mr J A Gill
Mr A C G Hayward
Mr I E Hunter
Mr D C Iles
Mr A Low
Mr S J Matthews
Mr B R Mawson
Mr J D Place
Mr W Ramsay
Mr K Ross
Mr R Thomas
Dr J Tubman
Mr P J Williams
Flint & N e i l l Partnership
Fairfield Mabey Ltd
Highways Agency
W S Atkins Consultants Ltd
Hyder, Special Structures
Cass Hayward & Partners
(formerly of Kvarner Cleveland Bridge Ltd)
The Steel Construction Institute
Ove Arup & Partners
WSP Consulting Engineers
Gwent Consultancy
Mott MacDonald
Corus Construction Centre
Railtrack
Rowecord Engineering Ltd
Scott Wilson Kirkpatrick
The British Constructional Steelwork Association Ltd
As a result of the representation of diverse interests in the Group, the Guidance Notes may
be considered to be guides to good, accepted practice. They should not, however, be taken
to provide advice that i s universally applicable to any and every project.
The SCI and the members of the SBG assume no responsibility for the adequacy of the
advice given, nor for the legal, contractual or financial consequences of i t s use.
Thanks are expressed to Corus (formerly British Steel) and to the Highways Agency for
financial support during the preparation of this publication. Additionally, SCI expresses
particular thanks for the contributions of the members of the SBG in drafting, debating and
revising the material for these Notes.
Comments, feedback and suggestions for further Notes will be welcomed; they
should be sent to SCI, for the attention of Mr D C Iles, at the address given on the
title page.
...
Ill
iv
CONTENTS
Section 1
Design - general
Section 2
Design - detailing
Section 3
Materials and products
Section 4
Contract documentation
Section 5
Fabrication
Section 6
Inspection and testing
Section 7
Erection and in situ construction work
Section 8
Protective treatment
Section 9
Other topics
USE OF THE PUBLICATION
The Guidance Notes are grouped in nine Sections.
subject, as listed under Contents above.
Each Section relates to one broad
Each Note i s assigned a three digit number, comprising the subject Section number and a
two digit serial number; for example, 2.04 i s the fourth Guidance Note in Section 2. Some
planned Guidance Notes are not yet issued, and the sequence therefore contains some gaps.
The number o f each Guidance Note i s given at the head o f every page. Each Note i s
separately page-numbered in a footer; the page number incorporates the Note number, for
example 1.01/2 i s the second page o f Note 1.01. Cross references from one Note to
another are denoted by prefixing the Note number by GN, for example GN 3.01.
A l i s t o f all the Guidance Notes in the first and second issues i s given overleaf.
V
LIST OF GUIDANCE NOTES
No.
Title
1.01
Glossary (revised)
1.02
Skew bridges
1.03
Bracing systems
1.04
Bridge articulation
1.05
(to be issued)
1.06
(to be issued)
1.07
Use of weather resistant steel
1.08
Box girder bridges
1.09
Comparison of bolted and welded splices
1.10
Half through bridges
2.01
Main girder make-up
2.02
Main girder connections
2.03
Bracing and cross beam connections
2.04
Bearing stiffeners (revised)
2.05
Intermediate transverse web stiff eners
2.06
Connections made with HSFG bolts
2.07
Welds - how t o specify
2.08
Attachment of bearings
3.01
St ruct u ra I st ee I s
3.02
Through thickness properties
3.03
Bridge bearings
3.04
Welding processes and consumables
3.05
Surface defects on steel materials
3.06
Internal defects in steel materials
3.07
Specifying steel material
4.01
Drawings
4.02
Weld procedure trials
4.03
Allowing for permanent deformations
5.01
Weld preparation
5.02
Post-weld dressing
5.03
Fabrication tolerances
5.04
Plate bending
vi
5.05
Marking of steelwork
5.06
Flame cutting of structural steel
5.07
Straightening and flattening
5.08
Hole sizes and positions for HSFG bolts
5.09
The prefabrication meeting
6.01
Weld quality and inspection
6.02
Surface inspection of welds
6.03
Sub-surface inspection of welds
6.04
Hydrogen/HAZ cracking and segregation cracking in welds
6.05
Weld defect acceptance levels
7.01
(to be issued)
7.02
Temperature effects during construction
7.03
Verticality of webs at supports
7.04
Trial erection and temporary erection
7.05
Installation of HSFG bolts
7.06
Transport of steelwork by road
7.07
(to be issued)
7.08
Method statements
8.01
Preparing for effective corrosion protection
8.02
Protective treatment of bolts
9.01
Construction (Design and Management) Regulations (revised)
vii
viii
SECTION 1
DESIGN - GENERAL
1.01
Glossary
1.02
Skew bridges
1.03
Bracing systems
1.04
Bridge articulation
1.07
Use of weather resistant steel
1.08
Box girder bridges
1.09
Comparison of bolted and welded splices
1.10
Half through bridges
Steel Bridge Group
Guidance Note
1.01
Glossary
Scope
This Guidance Note offers definitions of some
of the terms that are commonly used in the
steel bridge construction industry. Many terms
are common t o other forms of structural
steelwork, but some have a particular usage in
relation t o bridge construction (including terms
that have usage for bridges in materials other
than steel).
The definitions are provided by the Steel Bridge
Group t o assist readers of the Guidance Notes.
They should not be regarded as contractually
definitive.
Where definitions from other
sources are used (either in whole or in part),
the reference is noted at the end of the
definition; references t o definitions in BS 6 1 00
are given as a seven digit (3 + 4) number.
TERM
DEFINITION
abutment
An end support t o a bridge. It may or may not also act as a retaining wall.
beam
A structural member which carries loads principally in bending, and which spans
between main supports or between connections t o other members. The primary
elements of bridges are usually longitudinal beams.
bearing
A structural device located between the deck and an abutment or pier of the
bridge and transferring loads from the deck t o the abutment or pier (ENV 1993-2).
The bearing allows freedom of relative movement or rotation in some directions.
bearing
plate
A seating plate between the underside of a girder flange and the bearing that
supports the girder. See GN 2.04.
black bolt
A bolt in a clearance hole, designed t o transmit load in bearing/shear. No friction
between faying surfaces is taken into account, even if the bolt is pre-tensioned.
bracing
A secondary structural element of a bridge connecting t w o major parts, usually
for stabilisation of those parts against buckling, but also t o provide a load path for
forces transverse t o main beams, etc. (e.g. wind bracing).
brittle
fracture
The rapid propagation of a crack in a single application of loading, and without
any extensive plastic deformation.
camber
A design curvature in the vertical plane of a beam, girder or bridge deck. See
GN 4.03.
CEV
Carbon Equivalent Value. A representative measure of the alloy content of steel
that affects is weldability.
Charpy test A test t o determine a representative value of the toughness of steel material. A
small notched specimen is struck in a special testing machine and the energy
absorbed is measured (an impact test). See GN 3.01.
cope hole
A cut-out or ‘hole’ at the edge or corner of a planar element t o facilitate
connection t o other elements, e.g. at the corner of a web stiffener t o clear the
radiused fillet of a rolled beam section or at the weld of a fabricated beam.
crosshead
The ‘head’ of an intermediate support (pier), on which the bearings beneath the
main girders are seated. Alternatively, use is sometimes made of an integral
crosshead - a crossgirder (usually heavy) joining a number of main girders and
under which bearings are located (between the main girders and fewer in
number), each seated on a separate column.
deck
Strictly, the ‘floor’ of a bridge, but, for highway bridges, much more usually: The
superstructure of a bridge, comprising longitudinal beams, transverse beams or
bracing and either a reinforced concrete slab or a steel orthotropic plate (see
below) t o carry the traffic or services on the bridge.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.01/I
Guidance Note
1.01
TERM
DEFINITION
diaphragm
A transverse bulkhead inside a box girder, t o maintain its shape or to distribute
concentrated loads or support reactions.
erection
The activities performed to set in place on site the structural elements of a bridge.
erection
piece
A fabricated part of a girder that is a sub-division of a girder which is too long t o
fabrication
The activities performed t o assemble structural elements from plain materials.
faying
surfaces
The abutting faces between t w o parts of a connection. Usually applied t o the
interface of a HSFG connection that transmits shear force between the parts.
fitted
stiffener
A web stiffener that is prepared at its end and fitted t o the surface of the flange
that it abuts in such a way that the gap between the t w o is small and that loads
can therefore be considered t o be transmitted in direct bearing. See GN 2.04
flange
The upper or lower tension or compression element (usually planar) of a beam or
box girder. The corresponding element in an open web girder is known as a
chord; chords are formed of a rolled or fabricated sections.
flat
Rolled steel section of rectangular cross section whose thickness exceeds one
tenth of its width (210 5033)
gauge
length
A datum length over which various properties (e.g. elongation at fracture, out-ofstraightness, etc.) may be measured.
girder
An alternative term for a metal beam. Usually applied t o a beam that has t w o
flanges and a solid web. Girders may be rolled steel sections, but more commonly
the term is applied to a fabricated plate girder or t o a box girder. Truss girders
(see below) and vierendeel girders (see below) are 'open web girders'.
girder
make-up
The component parts that are connected together when a girder is fabricated.
Usually expressed in tabular form on drawings as a list of plate sizes and principal
secondary elements (intermediate stiffeners, stud connectors, etc.).
half joint
An in-line structural connection between t w o beams or parts of beams that is
formed by notching both, the upper on one side and the lower on the other. The
connection between the t w o allows rotation but not moment transfer; it transfers
shear and sometimes axial load.
be fabricated and erected in one piece. The length and weight of the erection
piece is chosen t o suit handling and transport.
half-through A bridge configuration in which the level of the road surface or track is below that
of the top flanges or chords (usually below mid-depth of the beams) and there is
no cross bracing between the upper flanges or chords. The bridge is usually a
square U in cross section. See GN 1.10. By contrast, a through bridge has
girders that are sufficiently deep that top bracing can be provided above the
traffic.
HSFG bolt
A nut and bolt assembly used in structural steelwork which relies for its effect on
the frictional force between the elements joined being developed by the clamping
action of the assembly. See GN 2.06.
ladder deck A configuration of t w o longitudinal main girders with cross girders spanning
orthogonally between them and the deck slab spanning longitudinally between the
cross beams. The steelwork arrangement is similar t o that of a ladder.
1.01I2
© 1998, 2000 The Steel Construction Institute
Guidance Note
1.01
TERM
DEFINITION
lamellar
tearing
The tearing of a steel element when subject to forces normal t o the plane of its
surface. See GN 3.02.
lamination
An elongated planar discontinuity in the material of a rolled plate or section.
location
plate
(In railway bridges: 1 A plate bolted t o the support structure upon which a bearing
is located and fixed; the plate distributes the load from the bearing t o the support
structure.
longitudinal Usually applied to: The direction of span of the bridge, or to: The orientation of
the greatest dimension in a structural element.
machined
surface
A surface that is ground, planed, milled or otherwise mechanically prepared to a
greater degree of flatness and smoothness than is obtained by rolling, flame
cutting, shearing, etc.
MPI
Magnetic Particle Inspection - a process for detecting surface breaking or
near-surface imperfections in steelwork, including welds, where the imperfection
is visually highlighted by the use of a strong magnetic field and iron-filings in a
carrier liquid.
normalised
Heat treatment of steel after rolling, or during rolling (normalised rolled). See
GN 3.01.
notch
A cut-out in an element. Particularly applied to a cut-out in a steel element when
the shape is such as to cause local stress concentration sufficient t o reduce the
fatigue performance or t o increase the susceptibility t o brittle fracture.
notch
A representative measure of the toughness of steel t o resist suddenly applied
load; measured in a Charpy test (see above). See GN 3.01
ductility/
toughness
orthotropic
plate
A plate with different stiffness properties in the longitudinal and transverse
directions. Usually applied t o a steel plate stiffened principally in one direction,
with widely spaced stiffeners or support in the other direction.
pier
An intermediate support t o a bridge deck. A pier may take the form of a wall (a
leaf pier) or a number of columns.
plate
Thin rigid flat metal product (210 5313)
Q and T
Quenched and tempered steel. See GN 3.01
steels
residual
stresses
Stresses locked into a steel element as a result of the production process. Such
stresses may arise from differential cooling/shrinkage rates during rolling
(particularly in rolled sections), from shrinkage of welds after deposition, or from
cold forming (bending plastically t o achieve a required shape).
shear
connector
A connecting device to transmit shear; the term is usually applied t o a connector
between a steel beam and a reinforced concrete slab. The most commonly used
form of connector is a forged headed shear stud (see BS 5400-5).
shrinkage
In steel: The reduction in dimension when steel cools after being heated.
In concrete: The reduction in dimension due t o decreased moisture content that
occurs after the concrete has been cast.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.01 /3
I
Guidance Note
1.01
TERM
DEFINITION
snipe
A cut-off corner (usually a 45" triangle cut off a square corner). Sometimes used
to form a cope hole (see above) at an internal corner.
splice
An in-line connection between t w o elements formed by overlapping, usually
involving secondary elements connected t o each primary element. The secondary
elements are termed splice plates or cover plates. Most commonly these plates
are bolted, but they may be welded.
stiffener
An element attached t o a planar steel element that increases its stiffness t o outof-plane deformation. A stiffener is usually linear in form, either of open section
(e.g. a flat, an angle or a Tee) or of a closed section (V- or U-shaped, connected
along both edges). Stiffeners may be either longitudinal or transverse t o the
element that they stiffen.
through
Deformation properties of the steel perpendicular t o the surface; these are often
thickness different from those in the direction of the plane of the surface. GN 3.02.
properties
TMR steels
Thermomechanically rolled steel. See GN 3.01.
tolerance
Permissible variation of the specified value of a quantity.
transverse
An orientation at right angles t o the longitudinal dimension of a structure or
structural element. (For a web plate in the vertical plane, transverse may be
taken as vertical if it is close t o being at right angles.)
trimmer
Usually applied to: local stiffening or support at the end of a reinforced concrete
deck slab, either by means of a transverse beam or by local thickening of the
slab. See GN 1.03. Also applied t o edge beams in heavily skewed bridges into
which beams square t o the abutment are connected.
truss girder A girder formed by a triangulated arrangement of chord members (in place of
flanges) and diagonal and (sometimes) vertical web members. Sometimes termed
an 'open web girder'.
UT
(or UTI)
Ultrasonic Testing (Inspection) - a process for detecting sub-surface imperfections
in steel materials, including welds, by the use of ultra-sound pulse generation and
response, usually with an associated cathode-ray display. See GN 6.03.
vierendeel
girder
A girder made from chord members (instead of flanges), with discrete web
members at intervals between the chords, square t o at least one of the chords.
weather
resistant
steel
Steel material with a chemical formulation such that it corrodes to produce a thin,
tightly adherent, impermeable, coating on exposed surfaces, obviating the need
for further protective treatment. See GN 1.07.
web
The (usually vertical) element of a beam or girder between the flanges (or chords)
that resists shear deformation of the beam.
web
stiffener
A structural element limiting the unsupported size of a web plate panel in a beam
or girder.
An intermediate web stiffener is a transverse stiffener used t o control out-of-plane
buckling of the web.
A bearing stiffener is used t o control buckling and t o form part of the load path
from the girder t o the bearing. See GN 2.04.
1.01/4
0 1998, 2000 The Steel Construction Institute
*
Guidance Note
1.01
TERM
DEFINITION
welding
A process of joining (steel) materials b y fusing the edgeslsides together b y applied
heating, usually w i t h additional filler metal and flux, under a protective gaseous
envelope.
For definitions of welding terms see BS 499- 1
wide flat
Steel plate that is rolled t o a specific width (up t o about 650 mm) so that it may
be used as a flange without any longitudinal cutting t o size.
References
1. The whole of BS 6100 is published as a
single hardback volume: Glossary of
building and civil engineering terms,
Blackwell, 1993.
2. ENV 1993-2: 1997, Eurocode 3: Design of
steel structures - Part 2: Steel bridges, CEN,
1997.
3. BS 499-1: 1991, Glossary for welding,
brazing and thermal cutting.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.0115
Steel Bridge Group
Guidance Note
No. 1.02
Skew bridges
Scope
This Guidance Note relates principally t o the
design and construction of deck-type and halfthrough skew bridge decks. Deck-type bridges,
which comprise steel girders supporting a
composite concrete slab at the top flange, are
most frequently used for highways, whilst halfthrough decks are commonly used for railway
bridges.
internal supports will usually be directly under
the main girders. Some typical arrangements
for deck-type multiple spans are shown in
Figure 1.
For the purposes of this Note, a skew bridge is
one where the longitudinal axis of the bridge
deck is not square t o the lines of its supporting
piers and/or abutments. In the Note, the skew
angle is taken as the angle between a line
square t o the supports and the longitudinal axis
of the bridge. Thus the greater the skew
angle, the higher (or more severe) the skew.
On single span deck-type bridges, the main
girders are usually arranged parallel to the
longitudinal axis of the bridge when either:
Curved skew bridges are not covered.
Main steelwork system
Deck-type construction is common for highway
bridges and is used for some railway bridges.
In continuous bridges of this type, the main
girders are usually arranged parallel t o the
longitudinal axis of the bridge, even for high
skews.
At the intermediate supports Of continuous
bridges, the girders can be supported either
directly, by individual bearings beneath each
girder, or indirectly, with the bearings providing
support to integral crossheads that frame into
the longitudinal girders. In 'ladder' decks,
intermediate support
4 bearings
deck axis
With continuous spans, variable depth or
haunched girders are generally best avoided
for skews over about 25O, because of the
geometrical complexity of the bracing.
the skew is less than 45º ; or
the deck is narrow in relation t o its span.
On a narrow single span bridge with a pair of
main girders spanning in the direction of the
bridge axis, consideration should be given t o
squaring up the ends of the deck. This is
particularly relevant for railway bridges, t o
avoid track twist.
However, when either:
l
l
the skew is more than 45"; or
the deck is wide in relation t o its span,
the main girders are more often arranged t o
span square t o the abutments. In the case of
a wide deck, the main girders will span directly
between the t w o abutments over the centre
intermediate support
2 bearings
(a) Skew 0< 25°
intermediate support
intermediate support
integral crosshead
(b) Skew 8 > 25"
Figure 1 Arrangement of girders for deck-type multiple skew spans
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.0211
Guidance Note
No. 1.02
that they span from the abutment t o an edge
girder. The main girders are usually framed
into these edge girders with a rigid moment
connection.
Typical arrangements for deck-type single
spans are shown in Figure 2.
Skew 0-< 45°
Skew 0
>
45°
Figure 2 Arrangement of girders for decktype single spans
Half - through bridges
For railway bridges in particular, the depth
available for construction may dictate that halfthrough composite construction is used, with
cross girders spanning between t w o main
girders located at the edge of the deck. The
cross girders may be partly or wholly encased
in a concrete slab. When a steel deck plate is
used, the transverse 'ribs' are arranged in the
same manner as described below for cross
girders.
For skews up t o 25°, the cross girders can be
either parallel t o the abutment (termed skewed
cross girders) or orthogonal t o the main girders.
When the cross girders in the centre part of the
span are arranged orthogonal t o the main
girders, at the end of the span the cross girders
nearest the abutment may be fanned t o the
trimmer beams (termed fanned cross girders).
For skews greater than 25°, the cross girders
are better arranged orthogonal t o the main
girders, and skew trimmer girders spanning
between the ends of the main girders are
provided at the abutments t o support the ends
of the cross girders (an arrangement termed
trimmed cross girders).
Typical arrangements for half-through railway
bridges are shown in Figure 3.
1.0212
Figure 3 Arrangement of girders for halfthrough railway bridges
Bracing in deck-type construction
In deck type construction, for skews up t o
about 25º, intermediate bracing can be either
parallel t o the line of the abutment bearings
(termed skew bracing), or orthogonal to the
main girders. Intermediate bracing will usually
only be needed t o brace girders together in
pairs rather than t o provide a continuous line of
bracing across the deck.
Skew bracing requires that the stiffeners t o
which it is connected are welded t o the webs
of the main girders at a skew rather than
square. There is no particular advantage in
making the bracing skew, except for link
bracing between t w o braced pairs of main
girders, where the bracings would otherwise be
staggered in plan. For skews over about 2 5 º ,
intermediate bracing is almost always arranged
orthogonally t o the main girders.
Irrespective of skew, bracing at the abutment
supports is usually best arranged parallel t o the
line through the bearings. This bracing will
also act as a trimmer beam t o the deck slab.
At the internal supports of continuous decks
where the skew is small (typically less than
© 2000 The Steel Construction Institute
Guidance Note
No. 1.02
line through the bearings. This bracing will
also act as a trimmer beam t o the deck slab.
A t the internal supports of continuous decks
where the skew is small (typically less than
25°1, the bracing is also usually arranged
parallel t o the line through the bearings. For
higher skews (over 25°) there are no hard and
fast rules, and the bracing for each individual
bridge should be tailored t o the particular
circumstances. Geometry, erection method
and deflections (including twist) must all be
considered.
Plate girder integral crossheads are often used
at internal supports. Integral crossheads are
usually arranged orthogonal to, and frame into
the longitudinal girders. They may either join
the main girders in pairs, or be continuous
between girders across the width of the deck.
Where the main girders are joined in pairs over
a single bearing at the internal support the
integral crosshead effectively acts as a bearing
support diaphragm. See GN 2.09 for guidance
on integral crossheads.
In multiple-girder bridges it may sometimes be
appropriate t o make the lateral system
continuous across all the main girders rather
than brace the girders in pairs. This is because
continuous bracing participates in the
transverse distribution of load, the more so
with increasing skew. As the skew increases,
cross girders of a size similar t o that of the
main longitudinal girders may be needed (see
Reference 1).
Vertical profile
Particular attention should be paid t o bridges
with a vertical profile that has t o follow a
vertical curve, as a different geometry is
required for each of the longitudinal girders.
This, together with any crossfall, will also lead
t o differing geometry in the elements of the
bracing system. For railway bridges there may
also be a requirement t o superimpose a live
load precamber on the vertical profile of the
girders.
Girder twist
A t the abutments or end supports of bridges
having high skews (typically 45" or more)
where the main girders are interconnected by
the deck or by intermediate bracings, the deck
will tend t o rotate about the line through the
bearings. This mode of rotation will also occur
during slab construction, once the main girders
have been interconnected b y bracing, and will
cause the main girders t o twist. This twist
effect is explained in GN 7.03. The resulting
out-of-verticality of the girders should be taken
into account in the design.
Note that
BS 5400-6 requires that the out-of-verticality
of the web at the supports is limited t o D/300
(minimum value 3 mm).
A pre-set t w i s t can be built into the girders so
that after deck concreting, the verticality of the
web will be within tolerance. The designer
should determine the pre-set t w i s t necessary
t o counteract the t w i s t that will occur. When
the bridge is constructed, this pre-set can be
achieved on site either by designing the
permanent bracings with appropriate geometry,
or in unbraced girders b y twisting them at the
support during erection. On the drawings, the
pre-set t w i s t should be illustrated as a rotation
angle of departure from vertical and given in
tabular form if t w i s t varies at different
locations.
The tendency of the deck t o rotate about the
line of the abutment bearings must be taken
into account in the choice o f the type of
bearings and their axes of rotation. With
elastomeric or pot type bearings, which allow
rotation about all axes, no special consideration
is required. However if linear roller or rocker
bearings are needed t o provide restraint against
lateral rotation of the girder, the alignment of
the axis of the bearing is an important
consideration.
With half-through railway bridges, the bearings
at the obtuse corners may need t o be set
relative t o those at the acute corners t o
facilitate the erection of the cross girders.
Twist also has implications for the temporary
works. In particular it is necessary t o consider
restraint t o the main girders t o prevent lateral
torsional buckling. Experience during erection
has shown that the actual movement that
takes place is frequently less than that
predicted by the designer.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.0213
Guidance Note
No. 1.02
Designers should consider every bridge with a
skew greater than 45° as a special case, and
determine the most appropriate method of
dealing with twist.
Deck slab
The design of the reinforcement in the deck
slab of a composite bridge should take into
account the local moments and vertical shear
forces in the slab and the longitudinal shear
between the slab and the girder as required by
BS 5400-5 Clause 6.3.3. Note that this clause
states that the spacing of bottom transverse
reinforcing bars should not be greater than
600 mm, and that 600 m m is also one of the
limits on the longitudinal spacing of the shear
connectors in BS 5400-5 Clause 5.3.3.
In deck slabs of composite bridges the most
efficient arrangement is generally for the
longitudinal reinforcement t o be parallel t o the
main girders, and the transverse reinforcement
t o be at right angles t o the main girders. In
skew decks and away from the edge of the
deck at the abutment, the slab effectively
spans square between girders, so the
arrangement with the transverse reinforcement
at right angles to the main girders is the most
efficient.
This does, however, lead t o
complicated detailing of the reinforcement at
the edge of the deck parallel t o the abutment.
Hence at small skews (less than 15°) it may be
preferable to place the transverse
reinforcement parallel t o the abutments. This
will result in little loss in efficiency (see
Reference 2).
triangles of slab without planks left at the edge
of the deck at the abutments are usually cast
using conventional formwork supported from
off the abutment. Alternatively, though less
commonly, precast 'specials' may be made t o
close these gaps.
Detailing skew stiffeners
Care needs t o be taken in specifying the size
of welds between webs and skew stiffeners,
as the dimension of the weld throat is
dependent on the angle between the web and
the stiffener. Reference should be made t o
BS 5135, Appendix A, Table 14. As skew
increases, the ability of the welder t o achieve
root penetration on the acute side of the
stiffener is impaired (see GN 2.05). Also with
highly skewed stiffeners, access for bolting the
bracing members needs t o be considered in
deriving the geometry of the stiffener.
References
1. Bridged in Steel 12, British Steel.
2. Concrete, Feb 1 9 6 8 and August 1968.
3. BS 5400-5: 1979, Steel concrete and
composite bridges. Code of Practice for
design of composite bridges.
4. BS 5135: 1984, Specification for arc
welding of carbon and carbon manganese
steels.
It is important that the spacing and positioning
of the shear studs on the girder top flanges
and the detailing of the transverse
reinforcement in the slab are co-ordinated so
as t o avoid unnecessary clashes with the
studs. Shear studs are usually detailed in rows
orthogonal to the axis of the girder, but if
skewed reinforcement is preferred, the studs
should be arranged on the skew t o suit.
Where precast plank type permanent formwork
is used for the slab, the planks are normally
placed at right angles t o the main girders. The
spacing and positioning of the shear studs and
the transverse reinforcement should be
co-ordinated with that of the planks. The
1.0214
© 2000 The Steel Construction Institute
Steel Bridge Group
Guidance Note
No. 1.08
Box girder bridges
Ref
Scope
This Guidance Note is intended t o give the
designer an insight into the use of steel box
girder bridges. The various types of box girder
bridge used for long and medium span highway
bridges are described. The use of box girders
for railway bridges and footbridges is also
covered.
General overview
Before the events leading up t o the Merrison
inquiry into the basis of design and method of
erection of steel box girder bridges (in the
1 9 7 0 s ) , steel box girders were commonly used
in the UK over a wide range of span lengths on
highway bridges. Following the introduction of
BS 5400-3 in the early 198Os, and w i t h the
advent of modern plate girder fabrication
methods, the use of box girders has become
much less frequent. Trends in highway bridge
design in recent years show decreasing use of
box girders, particularly for shorter spans.
For long span highway bridges, the choice of
bridge type will depend primarily on
economics. Notable long span steel box girder
bridges in the UK are:
Avonmouth
Cleddau
Friarton
Ref
1
2
Foyle
3
Scalpay
4
Main span
174 m
213 m
174 m
233 m
171 m
Opened
1974
1975
1977
1984
1998
These may be compared with the prestressed
concrete spans constructed in the UK Over the
same period:
Ref
-
Redheugh
Orwell
Skye
5
6
7
Main span
160 m
190 m
250 m
Opened
1983
1983
1992
A I Tweed
A47 Yare
10
11
M4 Neath
12
Main span
56 m
76 m
100 m
Opened
1983
1991
1994
These may be compared with box girder spans
over the same period:
Nene
M25 Poyle
Lower Lea
Ref
13
14
15
Main span
40 m
50 m
75 m
Opened
1984
1985
1991
T w o other box girder highway bridges of note
are the portal frame bridge with inclined steel
box legs over the River Tummel at Clunie and
the hybrid box/plate girder bridge over the
River W y e at Chepstow:
Clunie
Wye
Ref
16
17
Main span
67.5 m
93 m
Opened
1981
1988
Factors that determine the selection of a
medium span composite steel box girder bridge
are seldom purely economic.
The prime
consideration may be appearance (Nene, Lower
Lea). Other bridges ( M 2 5 Poyle, M 2 5 Gade
Valley, A 3 4 Newbury Bypass) are alternative
designs for structures originally conceived with
prestressed concrete box girder decks. In
these cases, t o ease acceptance of the
alternative, the external profile of the original
concrete design was maintained.
The torsional stability of the box girder may be
used to good effect where the highway
alignment is tightly curved, and also where the
mounting of services between the girders
precludes use of cross bracing between
girders. As a rough guide, for medium span
bridges the ratio of the cost of steel box
girders t o plate girders is between 1.5:l and
2:l.
Alternatively, cable stayed designs may be
preferred - examples in the UK are:
Ref
Kessock
Dee
8
9
Main span
240 m
194 m
Opened
1982
1997
In medium span highway bridges, the trend has
been for plate girder bridges t o span
increasingly greater lengths.
Over recent
years, in bridges w i t h continuous plate girders,
the longest spans have been:
Of the long span box girder bridges, Cleddau
and the main spans of Avonmouth and Foyle
have steel orthotropic decks. Friarton has a
composite lightweight concrete deck, and
Scalpay has a composite normal weight
concrete deck. The type of deck is determined
by an evaluation of weight saving against cost,
with orthotropic decks having the lightest
weight but highest cost for area of deck
surface. In medium span highway bridges,
composite concrete decks are the norm.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.0811
Guidance Note
No. 1.08
Deck cross section
For long span highway bridges, the box girder
cross section is usually a variable depth closed
rectangular section. The deep wide sections
can be fabricated in short lengths for erection
by cantilevering (Avonmouth, Friarton) or in
long lengths for 'big lift' erection (Foyle,
Scalpay). Cleddau has a single constant depth
trapezoidal deck girder that was erected in
cantilever with a 1 1 9 m long 'big lift' to close
the main span.
For medium span highway bridges the cross
section can be either a closed box or an open
top box (often referred t o as a bathtub
section). The section may be rectangular or
trapezoidal, and may be of variable or constant
depth. BS 5400-3, clause 9.1 6.2.1, states
that girders should be nominally symmetric, of
rectangular or trapezoidal cross section, and
with webs in single planes inclined at less than
4 5 ° from the vertical.
The designer will normally opt for a cross
section that keeps the box geometry, and
hence the fabrication, simple. If crossfall (or
superelevation) is constant along the length of
the bridge, rectangular boxes may be canted
so that the flanges are parallel to, and the
webs orthogonal to, the crossfall of the road.
Alternatively the webs may be kept vertical
and the bottom flange horizontal, with the top
flange following the crossfall. In this case the
webs are of different height and may also vary
in height along the length of the bridge t o
accommodate any variations in crossfall or
superelevation. Boxes that wind, and hence
have t o be built twisted, should be avoided.
Gade Valley is a rare example of asymmetric
boxes - all the boxes have vertical webs
except for the outer web of the edge boxes
which is inclined. This was done so that the
alternative steel design had an edge profile
matching that of the concrete box girder deck
of the original design.
-
-
the effective area of the flange, the saving in
material for shorter span bridges (up t o about
5 0 m) is generally considered t o be outweighed
by the increased complexity of the fabrication,
particularly for variable depth boxes.
Inclined webs may sometimes be used for
aesthetic reasons and, for example, when an
approach viaduct continues the ,shape of an
aerodynamic box used on the main suspension
or cable stayed bridge element of a crossing
(e.g. Beachley viaduct, Erskine).
Use of an open top section can realise savings
in the material in the top flange of the girder.
However this must be offset against the cost
of temporary (or permanent) diagonal bracing
at top flange level which has t o be provided t o
maintain the shape of the box during
fabrication and erection, and also its stability
under the wet concrete condition (see
Figure 2).
A constant depth rectangular cross section is
generally considered to be the easiest and thus
cheapest t o fabricate. A typical arrangement
is shown in Figure 1.
Although the use of inclined webs to make the
bottom flange narrower can improve the b/t
ratio of the bottom flange, and so maximise
1 .OB12
Additional costs are also incurred in forming
the soffit of the slab over the width of box,
© 2000 The Steel Construction Institute
Guidance Note
No. 1.08
and in cleaning and remedial painting after the
concreting and the stripping of the formwork.
In an open top box with inclined webs, the
span of the slab spanning transversely
between girders can be reduced, thus saving
on deck slab weight (the slab is optimally
250 mm thick, but frequently a variable
thickness, haunched over the girders, is used).
A typical arrangement is shown in Figure 3.
between 1 2 m and 3 9 m, and for both single
and double tracks. A typical example is the
replacement bridge at Clapham High Street
(Ref 18). See also GN 1.10.
Figure 4
Figure 3
Typical open-top box girder for
2-lane overbridge
Nowadays t w o main areas of use for box
girder decks are in railway bridges and
footbridges. Between them, by number, these
t w o areas account for more box girder bridges
than the highway bridge sector.
Railway bridges
Much current railway bridge work involves the
replacement of existing bridge decks with new
decks. The deck replacement may be for one
or more of a number of reasons:-
Changes in track configuration
Proposed changes in rail traffic
Repair of degradation uneconomic
Frequently only limited construction depth is
available, either due t o the required highway
clearance or because longitudinal timber type
construction is being replaced by ballasted
track t o simplify track maintenance.
In
addition the bridge construction has t o be
carried out in a very limited possession of the
track. With these constraints through-type
decks that can be prefabricated in large
sections and erected very quickly are
frequently preferred. A common solution is
Railtrack's standard 'Western Region' box
girder bridge design. This comprises t w o
trapezoidal boxes as outer main girders
supporting a steel deck with T beam cross
girders spanning transversely between the box
girders at bottom flange level (see Figure 4).
This is a generic design that provides for spans
Railway bridge using box girders
Footbridges
Footbridge decks are generally narrow and
have a relatively shallow span t o depth ratio.
They often have to be erected over busy roads
where only limited road closures are available.
In these cases prefabrication and fast erection
will usually determine the deck type. Much
useful guidance on deck types is given in
reference 19. A typical example of a box
girder deck is the Gablecross bridge in Swindon
(Ref 20).
Design
The design (and construction) of long span box
girder bridges is a specialist field and should be
undertaken by suitably experienced engineers.
The method of construction must be
considered integrally with the design so that
all loading conditions during erection are
analysed, and built-in stresses can be properly
aggregated.
For the design of medium span composite box
girder bridges, reference should be made t o the
SCI Design Guide (Ref 21).
Designers should be aware that the Health and
Safety Executive requirements relating t o entry
into confined spaces apply t o box girders used
in bridgework.
They apply both during
fabrication and during inspection and
maintenance of the bridge. The following
advice should be noted:
Small box girders should be detailed such
that welding does not require access from
the inside of closed sections.
Where openings for interior access are
provided, these have t o have minimum
dimensions of 460 mm by 4 1 0 mm or
4 6 0 mm if circular, although for easier
access a dimension of 675 mm is preferred.
For the size of box girders generally used in
SCI-P-185 Guidance Notes on best practice in steel bridge construction
1.0813
Guidance Note
No. 1.08
medium span highway bridges, openings of
this size through stiffened diaphragms are
unlikely t o comply with the requirements of
BS 5400-3, Clause 9.17.2.7 and will
therefore require a departure from
standards.
Where welding has t o be carried out inside
long girders during erection, openings for
ventilation should also be provided. To
minimise the hazard of working inside a box
girder at a welded site splice, the ventilation
opening should be near and the access
route from outside should be short.
Services
Carrying services inside box girders can appear
attractive, but consideration must be given to
safety in installation, inspection and the
maintenance of the services themselves.
Access for maintenance again raises health and
safety issues (see comments above).
Internal drainage
The accumulation of water within box girders
in service is not uncommon - increasing the
load and aggravating internal corrosion. For
this reason, taking deck drainage through a
box girder should be avoided, except possibly
for short joint-free sections.
Water will also collect inside box girders
through the phenomenon known as ‘breathing’.
Humid air that has entered the box during the
day cools at night and condensation forms on
the surface of the steel. Daytime warming is
insufficient t o re-evaporate the water, so the
volume of water builds up. This will occur
even in box girders that are nominally sealed.
For most bridges the remedy is t o provide
adequate drain holes (at least 25 m m diameter)
at the low points where the water would
otherwise pond, and t o detail transverse
stiffeners and other attachments to the bottom
flange with adequate cut-outs t o allow any
water to run down t o the drain hole.
The maintenance programme should provide
for regular inspection t o clear any dirt that
might block drain holes.
Protective treatment
The interior of a box girder is not normally a
severe environment and the protective
1.0814
treatment can be less than for exterior
surfaces.
Dehumidification and sealing
On larger bridges, an alternative (although not
normal practice in the (UK) to provision and
maintenance of internal protective coatings is
to provide an active dehumidification system
inside the box girders. A drainage system and
maintenance programme are needed for such
a system.
In smaller boxes, hygroscopic agents such as
silica gel may be employed, although these
need regular replacement t o remain effective.
Where the cross section is too small for safe
access, the boxes can be fully sealed. This is
often done in combination with internal
pressurisation and ‘fogging’ (a process where
a volatile corrosion inhibitor is introduced into
a sealed enclosure). If boxes are fully sealed,
the consequent pressure and temperature
effects must be taken into account in the
design. If fogged, airtight access hatches for
re-fogging should be provided.
References
1. Acier-Stahl-Steel, May 1977.
2. Structural Engineer, Dec 1980.
3. Proc. Instn. Civ. Engrs., May 1984.
4. Structural Engineer, Feb 1999.
5. Proc. Instn. Civ. Engrs., May 1984.
6. Proc. Instn. Civ. Engrs., Nov. 1983.
7. Proc. Instn. Civ. Engrs., Apr. 1987.
8. Proc. Instn. Civ. Engrs., May 1997.
9. Structural Engineer, Feb. 1999.
10. Building with Steel, Dec. 1981.
11. New Steel Construction, Feb. 1994.
12. Proc. Instn. Civ. Engrs., Aug. 1996.
13. Bridged in Steel 1, British Steel.
14. Proc. Instn. Civ. Engrs, Apr.1987.
15. New Steel Construction, Oct. 1994.
16. New Civil Engineer, 1 4 May 1981, and
Building with Steel, Dec. 1981.
17. Bridged in Steel 2, British Steel.
18. Bridged in Steel 7, British Steel.
19. The Design of Steel Footbridges, British
Steel.
20. Bridged in Steel 5, British Steel.
21. Design Guide for Composite Box Girder
Bridges, The Steel Construction Institute,
1994.
0 2000 The Steel Construction Institute
SECTION 2
DESIGN - DETAILING
2.01
Main girder make-up
2.02
Main girder connections
2.03
Bracing and cross beam connections
2.04
Bearing stiffeners
2.05
Intermediate transverse web stiffeners
2.06
Connections made with HSFG bolts
2.07
Welds
2.08
Attachment of bearings
- how to specify
Steel Bridge Group
Guidance Note
No. 2.01
Main girder make-up
Scope
This Guidance Note covers the make-up of
bridge girders. Whilst the Note is written
principally with Isection steel bridge girders in
mind, much of the information is equally
applicable t o steel box girders for short and
medium spans.
The selection of the configuration of the bridge
and its structural arrangement is part of conceptual design and is outside the scope of this
Note. Brief mention of typical arrangements is
made below, and some guidance on highway
bridge arrangements is given in Reference 1.
Bridge arrangement
The designer will determine the span lengths
of the bridge from consideration of the physical
dimensions of the obstacle t o be crossed, the
required clearance envelopes, the available
locations for the abutments and intermediate
supports (if more than one span) and aesthetics. For bridges of more than one span,
variable depth girders may be chosen, if appropriate.
The number of main girders in the deck cross
section is determined principally by the required width of the deck and the economics of
the steelwork. In composite highway bridge
construction the t w o arrangements that are
most commonly used are:
multiple longitudinal main girders with the
deck slab spanning transversely
a 'ladder deck' comprising t w o longitudinal
main girders with the deck slab spanning
longitudinally between cross beams.
For railway bridges, the requirement to minimise construction depth below track level may
necessitate the use of half-through construction, in which the deck is supported by cross
girders spanning between a pair of main girders. Typical examples are given in GN 1.10.
When there is no requirement to minimise
construction depth, the deck (of steel or
composite construction) may be supported on
longitudinal main girders. In this case, the
main girders are normally located directly under
each rail, to minimise transverse bending
effects.
Main girder make-up
Once the structural configuration of the bridge
and the number of girders in the deck cross
section has been selected, the designer will
determine the required sizes of top and bottom
flanges at the critical sections, namely those at
the positions of maximum moment. Similarly,
the maximum shear forces are critical for
determining the required thickness of the web
at the supports.
In order t o make the most economical use of
material, flange sizes and the web thickness
should be reduced away from the critical
sections (except for short spans, under about
25m). The make-up of the main girder is
determined by consideration of the available
sizes of steel plate of the required width and
thickness, the lengths of the spans and the
maximum length/weight that can be transported t o site and erected.
In the UK, the Corus (formerly British Steel)
plate products catalogue provides information
on the range of sizes that is available for
plates. Generally, plates of the thickness most
commonly used in bridge girders are available
in lengths up t o 18.3 m, although on some
occasions longer lengths may be available by
arrangement. Hence, the flanges and webs of
girders will normally (except for short single
spans) be fabricated from a number of plates
arranged end to end. Long girders must be
subdivided into a number of erection pieces for
handling and transport. Girder lengths up to
27.4 m can be transported by road without a
movement order (see GN 7.06), and fabricators
often transport longer lengths. This is also the
case for individual erection pieces.
The plates making up a girder or erection piece
may either be joined in the fabrication shop (a
shop splice), which usually means a full penetration transverse butt weld, or at site. A site
splice (sometimes known as a field splice) can
be either a bolted connection or a transverse
butt weld. (Note that some site splices are
made at ground level, to join several components before erection.)
Transverse butt welds are expensive so it may
be more economical t o continue a thick flange
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.0/1
Guidance Note
No. 2.01
sons. The width of the top flange may also be
influenced by its use as a walkway across the
bridge.
connectors. On large girders there may be
longitudinal stiffeners, and in box girders there
are also diaphragms and/or ring frames.
Flange thickness
Where the thickness of the flange varies, the
designer should keep the overall girder depth
constant and vary the web depth (see
Figure 2). Steps in the web can easily be
made when they are cut from plate. If the
web were constant depth, the step in girder
depth would be more difficult t o accommodate
in an automatic girder welding machine.
Web stiffeners
The size of any web stiffeners, their locations
and their orientation (e.g. whether transverse
stiffeners are square t o the top flange or truly
vertical) should be given on the make-up sheet.
See GN 2.04 and GN 2.05 for comment on
attachment of stiffeners.
Figure 2 Girder section
Site connected attachments
Details of other attachments, such as bracing,
diaphragms, etc. are usually given separately,
but the make-up sheet is often used t o give
references t o indicate which attachment will be
made where.
Doubler plates
Railtrack limits the flange plate thickness in
railway bridges to a maximum of 7 5 m m .
Thicker flanges are obtained by welding on
doubler plates; the doubler plate thickness is
generally slightly less than the primary flange
plate (the plate attached to the web). Doubler
plates are often curtailed short of the girder
ends. See GN 2.02 for comment on detailing
doubler plate details.
Web thickness
Whilst the thickness of web needed in midspan
regions is often quite low, a minimum
thickness for robustness should be observed.
A minimum of 10 m m is commonly adopted in
highway and railway bridges (and 8 mm, or
possibly even 6 mm, in footbridges). The
possible effects of erection conditions on thin
webs should also be considered carefully when
selecting the make-up (particularly where
girders are launched).
Attachments
The make-up of a girder includes all the pieces
that are attached to the girder in the
fabrication shop. The principal attachments to
plate girders are web stiffeners and shear
Shear connectors
In composite construction, the spacing of the
shear connectors is often varied along the
spans. The spacing and number in each group
should be given.
The starting point for
measurements should be clear (avoid appearing
t o locate the first group at the very end of the
girder), as should the line of measurement
(whether in horizontal plan or along the top of
the steelwork, for example).
Material grade
Whilst it is usually good practice not t o mix
grades of material in made-up girders (to avoid
risk of using the wrong grade), where there is
little risk of confusion, for example for thick
flange material, the selection of different
grades (e.g. sub-grade K2 rather than J2) is
normal.
If material with through thickness properties is
required in particular locations, this should be
noted. It is better not t o make this a general
requirement; only those locations that require
the enhanced properties should be so specified.
See GN 3.02 for guidance on through
thickness properties.
Information about make-up
Make-up information is usually communicated
in graphic and tabular form (e.g. typical cross
sections plus separate make-up sheets)
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.01 /3
Guidance Note
No. 2.01
T w o typical make-up sheets from actual
projects are included at the end of this
Guidance Note. Each includes many, though
not all, of the aspects mentioned above. They
should be considered as illustrative only, and
not taken as definitive.
Electronic methods of information transfer are
increasingly being used. Whilst this carries
many benefits, drawings are still of great
assistance in illustrating the designer's needs,
in review and in approving proposals made b y
the fabricator. Dialogue between designer and
fabricator is important t o make sure that all the
necessary allowances (for cutting and weld
shrinkage effects, allowance for permanent
deformation, etc.) have been made by the
appropriate party.
Geometrical information
It is common for geometrical information,
particularly relating t o the vertical profile, t o be
included o n drawings showing the girder makeup. This information is of particular relevance
t o the make-up of the web, since it completes
the information needed when ordering material
and cutting out the webs.
Note that any allowances for permanent
deformation need t o be associated w i t h the
vertical profile. See GN 4.03 for guidance on
allowances for permanent deformations.
References
1 . Design guide for composite highway
bridges, The Steel Construction Institute,
2000.
2. BS 5400-3: 2000, Steel concrete and
composite bridges. Code of Practice for design
of steel bridges.
Sample make-up sheets
The following fold-out sheets present typical examples of make-up sheets. The examples are:
(1 ) A 3-span composite railway bridge, with each track carried by t w o girders and with a concrete
slab acting compositely with the top flange.
(2) A 3-span haunched composite box girder highway bridge. There are four open-topped steel
boxes and the bridge is curved and tapered in plan.
2.01 14
© 2000 The Steel Construction Institute
Steel Bridge Group
Guidance Note
No. 2.02
Main girder connections
Scope
This Guidance Note, together with GN 2.03,
covers a number of the typical connection
details that occur in the fabrication and
erection of a bridge made from steel girders.
The details are representative of those that
have been used in practice, but are not the
only details that are suitable in all cases. Some
details are also appropriate t o box girders.
The main girder make-up determines where
connections within it are needed and also
affects the details that are required. Make-up
is covered in GN 2.01.
This Guidance Note covers the connections
within the main girder, including the
attachment of longitudinal web stiffeners.
Connections between the main girders and
bracing or crossbeams are covered in GN 2.03.
The design and detailing of bearing stiffeners
are covered in GN 2.04. The connection of
intermediate transverse web stiffeners t o the
main girders is covered, along with their design
aspects, in GN 2.05.
Detailing of bolted splices in main girders is
covered in GN 2.06. Guidance on h o w t o
specify welds is given in GN 2.07.
Guidance on the attachment of bearings is
covered in GN 2.08, and on the arrangement
and detailing of 'integral crossheads' in
GN 2.09.
and are more likely t o distort the flange
(creating a transverse curvature), as a result of
weld shrinkage.
Figure 1 Fillet welds to flange
Only the minimum required size for the weld
should be specified, except that a uniform size
should be used as far as possible along the
whole length of a girder.
Deep penetration fillet welds, as shown in
Figure 2, are usually achieved with the
submerged arc process using DC positive
polarity. They give a greater effective throat
than an ordinary fillet of the same measured
leg length. The designer should continue t o
specify weld size in the usual way, but the
actual leg length/throat size visible will be
slightly less. Because the visible weld size
does not indicate the full throat size for such
welds, inspection procedures for this type of
weld cannot rely on leg length measurements:
strict observance o f the weld procedure
specification is required.
Shop connections in main girder
Splices in webs and flanges
Shop welded splices in flange and web plates
will normally be full penetration b u t t welds. In
most cases, they will be made before the
pieces of the girder are put together.
Flange to web welds
The designer specifies on the drawings the size
of weld required between web and flange.
Where the shear forces are modest this is
often the minimum size that is practically
acceptable (6 mm). Fillet welds, rather than
butt welds, should be specified between web
See
and flange in almost all situations.
Figure 1. Butt welds require more preparation
0 2000 The Steel Construction Institute
Figure 2
Deep penetration fillet welds
Changes in flange thickness
As noted in GN 2.01, changes in flange
thickness will usually be made whilst keeping
2.0211
Guidance Note
No. 2.02
the overall girder depth constant and varying
the w e b depth.
Intentional steps in flange faces at joints need
t o be treated in the same way as unintentional
steps arising from distortions or rolling
margins. A t a butt welded joint, if the step is
greater than 15% of the thickness of the
thinner part, or more than 3mm, Clause
4.2.4.2 of BS 5400-6 requires the flange t o be
tapered at 1:4 (see Figure 3). A t a bolted
joint, the step should normally be all o n one
side (and appropriate thickness packs used),
unless the change of thickness is particularly
large.
1:4 taper
one face aligned through changes in w e b
thickness. Changes of thickness of up t o
3mm each side can be accommodated at a
butt weld without any further precautions, as
long as the centrelines are aligned. Greater
steps should be tapered at 1 :4 on each side.
If the w e b is spliced by bolting at a change of
thickness, steps of no greater than 1mm o n
each side can be accommodated without
make-up packs.
Site splices in main girder
Connections made on site will either use full
penetration butt welds or HSFG bolted joints.
Weld procedures for site welds are usually
similar t o those for shop welds. However,
because there will usually be greater limitations
o n weld positions (the pieces probably cannot
be turned) and environmental conditions may
be less favourable, the range of suitable
procedure may be more restricted.
'
Figure 3
Tapered change of flange thickness
Doubler plates
Where a very thick flange is needed, doubler
plates are sometimes used.
For railway
bridges, Railtrack limits flange plate thickness
t o 75mm, so a thicker flange will require a
doubler plate t o be provided. The doubler plate
is fillet welded onto the outer face of the
primary flange plate (the plate that is attached
t o the web), and should therefore be narrower,
t o allow room for the weld. Where doubler
plates are curtailed short of the girder ends,
they are tapered in plan, radiused around the
end, and tapered in elevation t o smooth the
stress f l o w (see Figure 4). Nevertheless, this
is still a Class G fatigue detail. (Note also that
the tapered portion is not included in the
effective section for stress analysis.)
Cope holes in webs
Where there is t o be a site weld across a
flange, or where there is a step in the web t o
suit a tapered change of flange thickness, a
semi-circular cope hole is usually provided. See
Figure 5. If the web is spliced at the same
position, the end of the weld (on the inside
face of the cope hole) should be ground flush.
Figure 5
Cope hole in web at a flange weld
Longitudinal w e b stiffeners
Longitudinal stiffeners are usually only needed
on the webs of I-beam girders when the girder
is deep. and the web is thin. This is most often
the case w i t h haunched girders where the web
is deeper at the pier positions.
Figure 4
Detail at end of a doubler plate
Changes in w e b thickness
Webs should normally be dimensioned centrally
on the flanges, without any attempt t o keep
2.0212
Stiffeners may be either flats or angles.
Angles are often used in regions of significant
compressive stress because they are efficient
in resisting buckling, but it is very difficult t o
© 2000 The Steel Construction Institute
Guidance Note
No. 2.02
apply, inspect and maintain protective
treatment t o the inside face of the angle.
Where angle stiffeners are used, they should
be turned w i t h one leg down, t o avoid trapping
water and debris.
Longitudinal stiffeners are normally attached by
‘all-round’ fillet welds. The welds on exposed
faces (i.e. other than inside a box girder) need
t o be continuous, rather than intermittent, and
on both faces, t o avoid potential corrosion
problems.
Continuous stiffeners
Longitudinal web stiffeners are usually
provided primarily t o enhance the shear
capacity of thin webs. To participate also in
carrying longitudinal stresses, the stiffeners
must be structurally continuous. They will
usually extend over several ’panels’ between
transverse web stiffeners.
Transverse stiffeners are usually notched t o
allow continuous longitudinal stiffeners t o pass
through (see Figure 61, except at bearing
stiffeners, where the longitudinal stiffeners
should be attached t o the faces of the bearing
stiffeners. Where a stiffener is notched, the
loss of section should be taken into account in
design.
Figure 6
Notches for longitudinal stiffener
constraint during fabrication, since the
longitudinal web stiffeners are usually attached
before the transverse stiffeners. One way t o
avoid this is by using ’over-width’ cut-outs in
the transverse stiffener and welding tongue
plates t o make the connection (see right hand
detail in Figure 6).
Welds between the longitudinal and transverse
stiffeners are normally needed only on the back
of the stiffener (BS 5400-3, clause 9.13.1,
specifies weld t o one-third of the perimeter).
Discontinuous stiffeners
If the longitudinal web stiffeners are only
needed t o stabilise the web they may be
discontinuous.
The use of discontinuous
stiffeners avoids the extra fabrication work in
attaching and welding the t w o types of
stiffener where they meet, but the detail at the
discontinuity needs t o be considered carefully
during design, particularly in regard t o fatigue
effects.
Where stiffeners are discontinuous, sufficient
clearances should be allowed for completing
welds and applying protective treatment. A
typical arrangement is shown in Figure 7.
References
BS 5400, Steel concrete and composite
bridges.
BS 5400-3: 2000, Code of Practice for
design of steel bridges
BS 5400-6: 1999, Specification for
materials and workmanship, steel.
The gap between the stiffener tip and the face
of the cutout is usually kept t o a minimum for
structural reasons, but it must be remembered
that protective treatment has t o be applied t o
the faces. A minimum gap of 1 2 m m (and not
less than 1.5 times the stiffener thickness)
should be provided, more if possible.
The direct attachment of the longitudinal
stiffener t o the transverse stiffener is a
0 2000 The Steel Construction Institute
2.0213
Steel Bridge Group
Guidance Note
No. 2.03
Bracing and cross beam connections
Scope
This Guidance Note covers the attachment of
bracing members and cross beams t o the main
Igirders, usually by means of connections to
stiffeners or cleats on the web of the girder.
The details are representative of details that
are used in practice, but are not the only
details that are suitable in all cases. Some of
the details described are also appropriate t o
box girders.
Additionally, the consequences of the use of
the various connections are covered in the
Note. The general considerations for design
and detailing of web stiffeners are covered in
GN 2.04 and GN 2.05.
Guidance on the arrangement and detailing of
'integral crossheads' is covered in GN 2.09.
For guidance on the design of bracing systems,
see GN 1.03.
Intermediate triangulated bracing
Where the main Igirders are deep enough,
diagonal bracing is normally used, either with
a single diagonal or crossed diagonals. See
Figure 1.
In detailing a connection, moments induced in
the stiffeners and bracing members can be
minimised by ensuring that:
the centrelines of bracing and effective
stiffener section meet as closely as is
practical
the distance between these intersections
and the web flange junction is minimised
(but sufficient clearance should be allowed
for application of protective treatment and
for access t o tighten the bolts).
The 'centreline' of a bracing member is,
strictly,
along its centroid, although for
triangulated bracing members the centrelines
indicated on drawings are often the lines of the
bolts.
Bracing members are usually attached b y
bolting. If the stiffener is sufficiently wide
they can connect through a simple lap
connection. Typical connections are shown in
Figure 3.
A 200mm wide flat stiffener
provides sufficient room for a two-bolt
connection using M24 bolts. The ends of
bracing members should be kept sufficiently
clear of the face of the web t o allow
completion of protective treatment. Similarly,
the horizontal leg of the bracing member
should be sufficiently clear of the flange t o
allow access for painting.
For shallower girders where diagonals would
be at too shallow an angle t o be efficient,
either K-bracing (see Figure 2) or 'channel
bracing' (see next page and Figure 5) is
normally used.
Figure 3
Lapped
member
connection
of bracing
Where t w o bracing members are bolted t o one
end of a stiffener, the t w o 'members should
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.03/1
Guidance Note
No. 2.03
generally be separately connected. It is not
good practice t o try t o 'economise' by lapping
t w o members on opposite faces of the
stiffener using a common group of bolts,
because it complicates erection.
member of the bracing acting as a tie, which
can generate significant moments at the top of
the stiffener, and this can give rise t o fatigue
problems. The weld size may need t o be
increased if the tie is left in permanently.
If there is insufficient room for the required
number of bolts within the width of the
stiffener (as chosen for its function as a
stiffener), the stiffener may be shaped and
extended t o facilitate the connection.
However, it is advisable not t o increase the
width over the full height of the stiffener, just
t o facilitate connections at the ends, because
an excessive outstand in the central region will
reduce its effectiveness as a stiffener. See
Figure 4.
Bracing can represent a greater maintenance
liability than the general steelwork. This is
particularly true of the top horizontal member,
which is more difficult t o maintain (because of
the slab above), is an excellent roosting spot
for birds and is also a dirt trap because of the
orientation of the angle. (Usually the horizontal
leg is at the bottom, t o avoid a significant
width of plate close t o the underside of the
deck slab.) Because of this and the potential
fatigue problem, horizontal top bracing
members should either be avoided or, where
used for construction purposes, removed
afterwards.
-
-
In choosing the arrangement of the bracing
itself, consideration should be given to the
assembly of the bracing and the space needed
for its attachment t o the web stiffeners.
Figure 4
Examples of shaped stiffeners
A t the crossover of X-bracing it is usual t o
provide a packing piece (the same thickness as
the web stiffener) and a single HSFG bolt
through all three pieces. This avoids a narrow
gap that is difficult t o maintain and achieves a
reduction in effective length in compression.
The designer should specify the geometry of
the bracing system and the number of bolts at
each connection. Either the designer or the
fabricator can detail the exact location of bolt
holes. Sufficient space should be allowed that
the fabricator has some flexibility in detailing
hole positions and accommodating tolerances
without infringing minimum edge distances.
Triangulated bracing provides 'torsional'
restraint t o the main girders.
The web
stiffeners t o which the bracing is attached
should therefore be welded t o both flanges
(fillet welds are adequate). Note, however,
that where there is a horizontal member close
t o the slab, it may attract loading as the deck
slab deflects under wheel loads. The rotation
of the girder top flange is restrained by the top
2.0312
Gusset plates
Gusset plates should always be lapped and
either fillet welded around the edges or bolted
t o bracing members, rather than, for example,
butt welding a smaller plate onto the heel or
toe of an angle. The latter detail is more
expensive t o weld and i t is very difficult t o
achieve a common plane surface between the
face of the bracing member and the extension
plate.
With K-bracing, the central gusset can be shop
welded t o the cross tie, leaving the diagonals
t o be bolted t o it during erection. The central
gusset will normally be rectangular (the
outstand corner may be sniped); there is no
saving in shaping it t o a V between the
diagonals.
Channel bracing
An alternative to triangulated bracing is t o use
a stiff cross member, usually a channel, to
provide torsional restraint t o the main girders.
A channel is usually chosen because, with
flange outstands on one side only, it is easy to
lap it onto the stiffener
© 2000 The Steel Construction Institute
Guidance Note
No. 2.03
Channel bracing requires a moment resisting
connection t o the girder (see Figure 5). This
is usually accomplished b y providing a group
of HSFG bolts. The web stiffener is normally
a simple flat, though it will probably be wider
than would be needed as an ordinary
intermediate stiffener. If it is over-width, it
should be tapered back t o the usual limiting
width either side of the connection t o the
channel.
If the channel bracing member is too shallow
for an adequate group of bolts, a plate can be
welded across the end of the member t o
provide room for further bolts. A simple full
depth plate lapped onto the channel is much
easier t o fabricate than angle or flat cleats
welded t o top and bottom, and achieves a
flatter surface for bolting t o the stiffener.
Plan bracing within the depth of the slab is
usually connected t o the girders b y means of
cleats welded on the top flange, although this
may complicate the positioning of shear
connectors and always complicates the fixing
of slab reinforcement. For these practical
reasons, bracing within the depth of the slab
is best avoided.
Connection of plan bracing below top flange
level is best achieved through cleats welded
into the corner between w e b and a transverse
stiffener (Figure 6 ) . Such bracing is usually
removed after the slab is cast, t o avoid the
need for future maintenance.
Plan bracing t o the bottom flange is not
normally needed, but if it is used, a detail
similar t o that in Figure 6 should be adequate.
When detailing the extra plate, there should be
space for fillet welds all round - the plate
should therefore extend slightly beyond the
end of the channel, not be flush with it. The
outer end of the plate may be flush with the
edge of the stiffener or project beyond it; it
should not be set back from the edge of the
stiffener.
Support bracing
Bracing at supports may be triangulated or
channel bracing. Similar considerations as for
intermediate bracing generally apply, though
the design loads and thus the member sizes are
larger. Additionally, this bracing will often
perform as a trimmer beam at the end of the
slab.
Figure 5
Connection of channel bracing
Plan bracing
Lateral restraint is sometimes provided by plan
bracing. Angle sections are usually adequate
for this purpose.
Plan bracing t o the top flange, if provided, is
usually required only before construction of the
deck slab. It can be located at mid-thickness
of the slab, but will preferably be connected
sufficiently below top flange level that it does
not interfere with formwork for the slab soffit.
Where the supports are at a significant skew
t o the line of the bridge, the deflection of the
beams and the bracing during construction
must be considered carefully. Effectively, the
bracing plane rotates about a line through all
the bearings, which results in a t w i s t of the
main girders about a longitudinal axis. The
designer must make it clear at which stage the
webs are required t o be vertical, as this will
determine the geometry of the bracing. See
GN 1.02.
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.0313
Guidance Note
No. 2.03
In some designs, a reinforced concrete
diaphragm is provided at end supports. This
reduces the amount of exposed steelwork in an
area where access for maintenance can be
difficult, and where water penetration through
deck joints can lead t o corrosion.
With
concrete diaphragms, light bracing may still be
needed for construction. Holes in the webs of
main girders for reinforcing bars should be
large enough to allow plenty of room for fixing
the reinforcement. Shear stud connectors are
often provided on the webs of the main girders
t o ensure that the concrete does not shrink
away from the face of the web.
Where channel bracing acts as a trimmer,
shear connectors are normally provided, t o
generate composite action of the trimmer
beam.
High skews lead t o difficulties for access (into
the acute corner) for welding the bearing
stiffeners. The size of the fillet weld in the
obtuse corner needs to be considered carefully.
See GN 2.04.
End connections in integral bridges
At the ends of integral bridges, the ends of the
steel girders will normally be cast into the
abutment (either capping beam or endscreen
wall detail).
As for reinforced concrete
diaphragms for non-integral bridges, light
bracing (which will probably get cast into the
abutment) will usually be necessary for the
construction condition. Holes are needed in
the web for reinforcing bars t o pass through,
and shear connectors will be needed t o assist
in transmitting the moments and forces. For
further advice, see Reference 1.
Connection of 'ladder deck' cross beams
The steelwork system in ladder decks generally
comprises t w o main longitudinal girders and
equally spaced cross beams arranged to span
transversely between the main girders (like the
rungs of a ladder). The cross beams are
usually arranged orthogonal t o the main
girders, but for small skews ( < 1 5 ° ) can be
parallel t o the line of the supports. The cross
beams will usually be fabricated Isections,
although in some instances it may be possible
t o use a rolled section. The concrete deck slab
acts compositely with the main girders and the
2.0314
cross beams, over much of the deck effectively
spanning between the cross beams. There are
shear studs on the top flanges of the main
girders and the cross beams.
With a deck slab of uniform thickness, the top
flange of the cross beams will line up with the
top flange of the main girders. If the slab is
haunched over the main girders, the top flange
of the cross beams will be set up higher than
that of the main girders. This is often done t o
provide extra depth at the root of the slab's
edge cantilever if the cantilever is long.
Usually on ladder decks the cross beams are
provided only between the main girders,
however they may be extended as cantilevers
on the outside face of the main girders. This
allows the steelwork t o support the entire deck
slab in the wet concrete condition. Note that
BS 5400-3 clause 9.15.1.3 requires that any
transverse cantilever member is continuous
with a transverse member between the main
girder webs.
There are t w o methods in general use for
connecting the cross beams t o the main
girders. With a lapped connection, the web of
the cross beam is lapped on one side of a
stiffener on the main girder as shown in
Figure 7. With a splice plate connection, the
web of the cross beam is aligned with a
stiffener on the main girder and cover plates
are used as shown in Figure 8. Use of end
plate connections where an end plate welded
onto the cross beam is bolted t o the web of
the main girder is not recommended, because
of the difficulty in fit-up.
Figure 7
Cross beam lapped connection
© 2000 The Steel Construction Institute
Guidance Note
No. 2.03
The connections are usually designed so that
only a bolted connection t o the cross beam
web is required. This greatly simplifies the
steelwork erection as it is not necessary t o
make any connection between the cross beam
top flange and the main girder top flange, or
the cross beam bottom flange and the main
girder web stiffener. In the bare steel
condition, the connection must have adequate
capacity. for the w e t concrete loads, after
which the composite top flange will provide
additional capacity for subsequent stages of
loading. If need be, the web of the cross beam
can be strengthened b y welding on additional
plating in the region of the bolt group.
References
1. Steel integral bridges: Design guidance,
The Steel Construction Institute, 1997.
2. The Structural Engineer, Vol. 67 No. 1 5 / 1
August 1989.
3. The Structural Engineer, Vol. 68 No. 6 / 2 0
March 1990.
Further reading
Hayward and Weare, Steel detailer’s manual,
BSP, 1989.
Tordoff D, Steel bridges, BCSA, 1985 (due t o
be revised and reissued in 2000).
A t internal supports of continuous multi-span
bridges, the standard cross beam can be
adapted t o provide restraint t o the main girder
bottom flange at the support and, if needed,
within the hogging moment region. This can
be achieved by the introduction of bracing
members or b y increasing the depth of the
cross beam (see References 2 and 3). Similar
arrangements can be used at the abutments or
end supports. A t internal supports t o skew
decks, with judicious spacing of the cross
beams, the bearings under the main girders can
be arranged so that they are staggered by a
multiple of the cross beam spacing. A t a skew
abutment the cross beams can be framed into
a heavier end trimmer girder that is parallel to
the line of the bearings.
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.03/5
Steel Bridge Group
Guidance Note
2.04
Bearing stiffeners
Scope
This Guidance Note gives information about
practical detailing for bearing stiffeners for
plate girder bridges. The support positions of
box girders usually require plated diaphragms;
diaphragms and their stiffeners are outside the
scope of this Note.
General
This Note describes first the general design
requirements for stiffeners, then explains the
practicalities of certain details and finally
shows what the consequences of these practical considerations are on the design details.
Design requirements
BS 5400-3, clause 9.14.1, recommends that
the webs of plate girders and rolled beams be
provided with a system of load bearing stiffeners at each support position, and advises that
the section of a bearing stiffener should be
symmetrical about the mid-plane of the web.
When this condition is not met, the designer
must take into account the effect of the resulting eccentricity. (But note that 'true' symmet r y is not necessary; provided that the centroid
of the effective stiffener section lies in the
web, the condition is effectively satisfied.)
plate on each side of the stiffener. The axial
force in the stiffener will be at a maximum
immediately above the bearing and is dissipated over the length of the stiffener as load
is shed into the web.
Clause 9.14.1 (as recently revised) advises
that the ends of a bearing stiffener should be
"adequately connected t o both flanges but
only advises that it be "closely fitted" t o the
flange subject t o concentrated load.
The requirement that the bottom edge of the
bearing stiffener should be closely fitted t o the
bottom flange is so that the transfer of load at
the ultimate limit state may be assumed t o be
in direct bearing between the parent metal of
the stiffener and the upper surface of the
bottom flange. If the web of a fabricated
girder is required t o be included in the bearing
area, it should also be fitted locally.
As there is no residual axial force in the stiffener where its top edge meets the underside
of the top flange, there is usually no need for
the top edge of the stiffener t o be closely fitted
t o the flange.
Edges which are required t o be fitted should be
clearly indicated on the drawings.
Figure 1 Arrangement at a support bearing
The usual support arrangement is for the
bridge girder t o be placed with the bottom
flange seated on top of the bearing (see
Figure 1). A tapered bearing plate, sometimes
known as either a 'sole plate' or a 'seating
plate', is nearly always required between the
upper surface of the bearing and the underside
of the flange so that the bearing remains truly
horizontal. With this arrangement the load path
for the transfer of the bearing reaction t o the
girder is through the thickness of the bottom
flange plate and into the effective section of
the stiffener acting with a portion of the web
As well as being required t o carry the axial
force from the bearing reaction, the stiffener
must also resist bending moments arising from
eccentricity of the axial force (relative t o the
centroid of the stiffener section), flexure of the
end frame, and horizontal loading (FR forces,
which control buckling of the main beam, and
forces due t o wind) acting transversely t o the
girder. These are identified in BS 5400-3
clause 9.14.3. It is in order t o resist these
forces that the stiffener should be adequately
connected t o the top flange (as well as t o the
bottom), although this does not mean that it
has t o be fitted. Attachment t o the top flange
of a deck-type bridge also provides torsional
restraint t o the flange plate (see comment in
GN 2.05).
Eccentricity of the axial force can arise from a
number of causes, including thermal expansion/ contraction, bearing misalignment, beam
rotation under load, substructure movement
and shrinkage/creep of composite decks.
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.04/1
Guidance Note
2.04
On long continuous decks, the range of movement remote from the fixed bearing can be
significant. In such a case, the eccentricity of
the bearing reaction t o the stiffener centroid
can lead to large bending moments in the
effective stiffener section. This is usually
accommodated either by using multi-leg stiffeners, or by mounting the bearing 'inverted' (if
of the sliding type) so that the bearing remains
fixed in position relative to the superstructure
steelwork. The movement is then relative t o
the substructure, which can more easily accommodate eccentric loading. However, when
the bearing is mounted inverted, the sliding
surface will be facing upwards instead of
downwards, and it is therefore desirable t o
specify that the bearings are shrouded t o
prevent the ingress of harmful debris.
Between the end of the Stiffener and the flange
before welding, and this is a normal feature of
the fabrication process. Where the stiffener is
fitted, the majority of the load is transmitted in
direct bearing, so a full penetration butt weld
is unlikely to be necessary for ULS loads.
Multi-leg stiffeners may also be necessary on
haunched girders where there is a change of
direction of the bottom flange either side of the
bearing. It is advisable t o design the connection between such stiffeners and the flange t o
resist the tensile force due t o the change of
direction of the flange plate, ignoring the
bearing reaction. This will cater for situations
where the load path from the bearing does not
spread t o the stiffener as a result of extreme
thermal movement and during jacking for
bearing replacement.
Clause 6.3 of BS 5400-10 recommends that
none of the fatigue loads be carried in bearing,
so the fatigue checks must be based on weld
throat area. This may lead designers t o choose
a butt weld detail in preference t o fillet welds.
However, it is advantageous t o avoid a full
penetration butt weld, because shrinkage
strains (vertically) will distort the transverse
profile of the flange (see Figure 2) and cause
misfit of the bearing. Where the bearing
stiffener is thick, it may be helpful t o specify
a partial penetration weld, as this will reduce
the area over which fitting is t o be achieved
and the amount of weld metal t o be deposited.
Practicalities of fabrication
For the designer, the main consideration is the
type of welds that should be specified at the
three edges of the stiffener that are connected
t o the girder. These connections are:
stiffener to web
stiffener t o flange adjacent t o the bearing
stiffener t o flange remote from the bearing.
The connection t o the web will normally bethe
first edge welded. Fillet welds should be
sufficient in most Cases. Full penetration butt
welds are unlikely t o be necessary and are
undesirable; the cruciform detail (stiffeners on
either side of the web) creates restraint during
welding and can result in lamellar tearing.
The connection to the bottom flange is
normally the most heavily loaded part of the
stiffener. The stiffener can be 'fitted' t o the
flange such that there is very little gap
2.0412
BS 5400-6, clause 4.3.2 requires that the ends
of stiffeners that are required t o be fitted shall
be ground where necessary so that the maximum gap over 6 0 % of the contact area does
not exceed 0.25 mm. However, the clause is
not clear about the fit of the unmentioned 4 0 %
of the contact area. Clause 4.301 in the
Model Appendix 18/1 (Ref 2) is intended t o
remove this uncertainty by calling for a maximum gap of 0.25 m m over 6 0 % of the contact
area and 0.75 m m over the remainder.
Before welding
Figure 2
After butt welding
Distortion due to butt welding
(exaggerated)
To provide proper fitting at the stiffener t o top
flange connection as well as at the bottom
flange is difficult, because it requires a tight fit
between t w o nominally parallel faces. A fillet
weld connection to the top flange will normally
be adequate structurally. Where gaps occur at
the top flange as a result of fitting t o the
bottom flange, the size of the fillet welds may
need to be increased (see Figure 3).
If bearing stiffeners are skew t o the web, it
may be more difficult t o fit them t o the bottom
© 1998, 2000 The Steel Construction Institute
Guidance Note
2.04
I
flange, because plates may have a small ripple
due t o rolling; this ripple is too small t o have
any structural significance but may cause
problems in achieving the close fit if the flange
is
Figure 3
Connection of a bearing stiffener
Cope holes may be used t o clear root fillets or
web/flange welds, as for intermediate stiffeners (see GN 2.05), although this may detract
significantly from the bearing area at the
bottom flange and is best avoided there.
I
defects and t o repair. As an approximate
guide, the clear line of sight t o the root of a
fillet weld should make an angle t o the face of
the web of preferably no more than 30°, and
in all cases no more than 35" . See Figure 4.
This would allow access for a gloved hand
holding an electrode and a line of sight for a
welder wearing a mask. For a partial penetration weld the line of sight should be no more
than 20" t o the web (but better still, avoid
in these
p a r t i a l penetration
welds
circumstances).
Skew stiffeners also complicate the welding
detail at the connection t o the web. If the
skew is any more than 30° the fillet in the
acute angle (which is then less than 60°) will
not achieve full root penetration. Advice in
BS 5 1 35, Appendix A, says that, for fillet
welds, the full strength should not be relied
upon where the fusion faces are at more than
120° or less than 60°. The Appendix also
gives factors (less than the usual 0.7) for
deriving design throat thickness from leg length
when the angle is between 90° and 120°; this
necessary reduction is sometimes overlooked
by designers. Specifying the weld throat,
rather than the leg length, is a better way of
ensuring that sufficient weld is provided.
When the bearing reaction is very eccentric
from the web (for example above a knuckle
bearing which is providing torsional restraint to
the main girder), Tee stiffeners may be specified, but this does add considerably t o the task
of welding and t o the difficulty of the application of protective treatment.
Where there are multiple stiffeners, it is inevitable that some pieces will have t o be welded
once other pieces have been attached. To
ensure adequate welding, there must be access
for the operative t o lay the weld, access for
visual inspection and NDT, and, in the case of
unacceptable defects, access t o remove the
Figure 4
UOUbIe leg bearing stittener
Where stiffeners of the Closed-box type are
specified, they may either be pre-assembled
before fitting t o the web, or they may be built
up piece b y piece. In either case, it is not
possible t o lay welds at every corner on each
face, because closed spaces are created. The
detail chosen by the designer will effectively
determine the sequence in which the box must
be built up; the constraints inherent in the
implied sequence should be considered carefully at the design stage. A typical simple
configuration is shown in Figure 5; in this
example, the t w o outstands are welded first
and then the closing plate is fillet welded t o the
tips of the outstands-
Figure 5
Attachment of typical box-section
stiffener
Support at bearings
Careful consideration has t o be given t o the
load path between the girder and the bearing
itself. There are normally t w o interfaces t o be
considered:
a) bottom flange to bearing plate, and
b) bearing plate to bearing.
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.0413
Guidance Note
2.04
The top plates of proprietary bearings are
usually machined t o the tolerance given in
BS 5400-6, clause 4.2.3. The design of the
bearing normally assumes that load is applied
t o the top evenly or (for bearings resisting
rotation) in a linearly varying fashion, i.e. there
are no stress concentrations. The bottom of
the bearing plate should therefore also be
machined t o clause 4.2.3.
The other interface, between bottom flange
and bearing plate, needs careful consideration.
The top of the bearing plate is often machined,
because it is tapered t o suit an inclined soffit
and horizontal bearing top surface. However,
if the bottom flange were machined before
fabrication, that would not guarantee a surface
within the limits of clause 4.2.3, because of
subsequent distortions due to welding. On the
other hand, machining the bottom flange of a
completed girder is impractical. In any case,
machining the flange removes flange thickness
in a highly stressed region and this is likely t o
incur the penalty of a thicker flange t o allow
for the loss.
If a good fit is not achieved between the
bearing plate and the bottom flange, the load
will be transferred unevenly. However, it may
be possible t o show that, no matter where the
load is transferred at this interface, the 6 0 "
spread of load* through the plates on either
side of the interface is sufficient t o ensure that
the bearing and the web or stiffener forming
part of the effective bearing stiffener are not
overstressed. However, in many situations
this will not be the case, so this interface will
often need t o achieve a tolerance equivalent t o
a machined surface; this can be achieved by
grinding the bearing plate or bottom flange of
the girder (see Model Appendix 18/1,
clause 4.2041, t o ensure even bearing. This
tolerance only has to be achieved over a
cruciform area equal t o the width of stiffeners
or web plus 60° spread through bottom flange.
Through-thickness performance
Where large transverse loads are carried
through the web (usually as a result of
diaphragm action or crosshead details), the
web may have to be fabricated with material
with guaranteed through thickness properties.
See GN 2.09 and GN 3.02.
Summary for design
Fillet or partial penetration butt welds, rather
than full penetration butt welds, should be
specified for the stiffener t o bottom flange
connection. For design t o BS 5400, the
designer should ensure that the following are
adequate:
yield strength of full effective section to
carry full axial load plus moment
buckling strength of full effective section,
t o carry the maximum load in the middle
third of the stiffener
bearing strength (at 1.33 times yield) to
carry full axial load and moment on the
section within spread lines* from the
bearing over that area of web and stiffeners
that has been fitted or butt welded t o the
flange
fatigue life for the stress range due t o loads
carried through welds between flange and
stiffener, based on the section determined
by spread from the bearing' and ignoring
direct bearing
strength of welds between stiffener and
girder web, t o transfer the loads between
the stiffener and the web
strength of the welds between the stiffener
and the top flange, t o carry lateral forces
(restraint forces, wind forces, etc.).
References
1. BD 13/90, Design of steel bridges: use of
BS 5400: Part 3, Department of Transport,
1990.
2. Specification for structural steelwork for
bridges: A model appendix 1811, The Steel
Construction Institute, 1996.
3. BS 5400-3: 2000, Code of Practice for
design of steel bridges.
4. BS 5400-6: 1999, Specification for
materials and workmanship, steel.
5. BS 5400-10: 1980, Code of Practice for
fatigue.
* The spread from the bearing, through the
bearing plate and the flange, is a t 60° to the
vertical, as given by BS 5400-3, Figure 28.
2.0414
© 1998, 2000 The Steel Construction Institute
'
Steel Bridge Group
Guidance Note
No. 2.05
Intermediate transverse w e b stiffeners
Scope
This Guidance Note gives information about
the need for intermediate transverse web
stiffeners for plate girder bridges and summarises the practical detailing aspects. Further
guidance on detailing is given in GN 2.02.
Stiffeners used as part of a U-frame in halfthrough construction require additional considerations that are not covered in this Note.
This Note first describes the main purpose of
intermediate transverse web stiffeners, then
describes their secondary role in the context of
intermediate restraints. It then explains the
practicalities of certain details and the influence of these details on the design and cost of
fabrication.
Primary purpose of intermediate web stiffeners
Where Universal Beam sections provide adequate bending resistance, they could, in theory, normally be used without any intermediate
web stiffeners (in the absence of any need for
intermediate systems of restraint to compression flanges). This reflects the inherent stability of webs in such sections and the general
ability of the section t o sustain significant
levels of plastic rotation before attainment of
ultimate load (i.e. the section classification is
generally compact).
The t w o key factors determining the stability
and hence load carrying capacity of webs are:
web slenderness (the d we/ tw ratio); and the
aspect ratio of web panels (the a/dwe ratio) see Figure 1. Web capacity reduces as these
ties in shear and bending; the overall capacity
of the girder would therefore be reduced.
From the above, the primary role of intermediate web stiffeners is clear: they serve t o
divide the web into panels and thereby reduce
the web panel aspect ratio. This increases
web capacity.
The maximum capacity achievable in a web
occurs if its geometric characteristics (d we/t w,
a/d we) are such as t o permit the development
of tension field action. This is an ultimate
condition in which a web panel buckles across
one diagonal, a very strong tensile band develops along the other, and the final load carrying
system essentially relies on truss action as
shown in Figure 2. In this situation, the stiffener and an associated vertical band of web
form a truss vertical 'post' and must be capable of sustaining the truss loads.
On deck-type bridges, intermediate stiffeners
are usually avoided on the exposed face of the
outer girders for aesthetic reasons. For halfthrough railway bridges, the intemediate
stiffeners are located on the outer face of the
girder in order that the gap under the top
flange on track side can, if necessary, be used
as a refugeSecondary roles of intermediate web stiffeners
In addition to their role in determining web
panel sizes and being active in the development of tension field action, intermediate web
stiffeners are frequently required t o play t w o
further roles:
To permit attachment of bracing between
beamd/girders.
Plate girders are generally different from
Universal Beams in relation to web stability.
The designer's freedom t o proportion the
component plates of such girders frequently
means that , ratios can be relatively high.
This, coupled with high aspect ratios (a/dwe
for web panels would result in low web capaci-
To act as load bearing stiffeners t o the web
at points of stationary concentrated loading
on the beam/girder (but not commonly in
bridges).
The first of these roles is discussed in the
context of detailing below. The second is an
2.05/1
SCI-P-185 Guidance Notes on best practice in steel bridge construction
Guidance Note
No. 2.05
aspect that is taken into account in the design
procedure of the strength and stability checks
in BS 5400-3.
Economy in design
The amount of workmanship in stiffener assemblies is disproportionately high compared
t o the other components in a typical bridge
girder. The cost of stiffening work varies
considerably among fabricators, depending on
their particular investment strategies and the
processes used. However it is often the case
that selecting a thicker web and reducing the
number of stiffeners will lead t o a cheaper
girder.
Care needs t o be exercised in relation t o the
proportioning of webs and stiffeners if the
most economical girder configuration is t o be
achieved.
Stiffener layout and detailing
Simple intermediate web stif feners
A typical arrangement for a simple single
intermediate web stiffener is a flat welded onto
one face of the web and connected t o one
flange (see Figure 3).
Figure 3
Simple web stiffener
Flat stiffeners are usually preferred as they are
are easily welded onto web plates using allround fillet welds. They are normally attached
after the web has been joined t o the flanges
(with automatic girder welding they have t o be
attached afterwards, since they would obstruct
the passage of the welding heads).
No more stiffeners should be specified than are
actually necessary for structural adequacy - by
any method of fabrication they are relatively
expensive items.
Flat stiffener outstands (width of the flat) are
limited by BS 5400-3 (to 10t for S355 steel).
Sizes should be specified that are standard
rolled flat sizes; commonly used sizes include:
1 0 0 x 1 0 or 1 2 m m
150xl5mm
200 x 2 0 m m
2 5 0 x 25 m m
A t the top flange, attachment of the stiffener
t o the flange is not strictly necessary for its
function as a web stiffener. Where the top
flange is composite, the stiffener should be
attached, t o provide torsional restraint to the
flange. Fillet welds are generally sufficient for
attachment.
Only the minimum weld required for strength
should be specified between any web stiffener
and the flange (subject t o a practical minimum
size of 6 mm). Even fillet welds introduce
some local distortion of the flange, though
much less than would be introduced by a butt
weld. However, with a thick flange (t >
5 0 mm) it may be the choice of the fabricator
t o lay a bigger weld than specified t o avoid the
necessity for local preheat, which can itself
lead t o distortion.
At the bottom flange, a clear gap should be
left, preferably about 3 times the web thickness, but not more than 5t. This accommodates tolerances in depth of both the web and
the stiffener, whilst not being sufficient t o lead
t o fatigue problems at the stiffener end (except
Possibly where the web relies heavily on
tension field action - it is better t o attach both
ends in such cases). The free corner of the
stiffener is often scarfed, which facilitates the
application of protective treatment, as shown
in Figure 4.
Traditionally, web stiffeners have had t o be
attached and welded manually, though robotic
equipment is now becoming available that can
weld them automatically after they have been
positioned and tacked manually.
2.05/2
© 2000 The Steel Construction Institute
Guidance Note
2.05
Sniped corners (45") cut across the internal
corner) have been used by some designers in
this location, but the detail should be avoided,
because it is very difficult to return the weld
(i.e. in the acute angle, across the thickness of
the stiffener) and protective treatment cannot
be applied satisfactorily to the inner edge of
the stiffener or the weld return.
Note. Scarf angle typically 2:l
Figure 4
Detail at stiffener bottom
The shape of web stiffeners attached t o a
flange must allow for the fillet weld between
web and flange. One way of achieving this is
b y use of 'cope holes'. These are circular
quadrants cut of the right angle t o leave a
clear gap (see Figure 5). A radius of at least
40 mm, preferably 50 mm, is needed t o allow
welds t o be returned around the corners and
Protective treatment t o be applied t o the inner
faces of the hole (although the standard of
cleaning can never be as good as o n a plane
The weld attaching an intermediate or other
web stiffener (i.e. across the flange and around
the edge) should n o t have its toe closer than
10 m m t o the edge of the flange, or a lower
class of fatigue detail will result. It is wise t o
choose a stiffener width at least 25 m m less
than the nominal flange out stand t o ensure
that this clearance is achieved in practice. If
it is necessary t o use wider stiffeners, perhaps
t o suit bracing, they can be scarfed at the
ends, or a quadrant notch (similar t o the cope
hole detail) could be made, as shown in
Guidance Note
No. 2.05
Stiffeners for ‘channel’ bracing
Again, flat stiffeners are preferred for attaching
channel bracing, though they will probably be
more substantial than those used with
triangulated bracing, since a moment restraint
has to be provided and a larger group of bolts
is needed to transmit the moment. Stiffeners
are sometimes shaped to provide extra space
for the bolt group. See GN 2.03.
Treatment of skew
Small degrees of skew of bracing members can
easily be accommodated by suitable alignment
of the flat stiffener, but it is normally just as
easy, and cheaper, to detail orthogonal
bracing. Highly skewed bracing should always
be avoided, because it introduces difficulties in
fabrication as well as being less effective as
intermediate bracing.
See GN 1.02 for
guidance on skew bridges.
Summary of web stiffener layouts
The layout of common arrangements of intermediate web stiffeners, and comments on
detailing, are summarised in Table 1. The
Table indicates the suitability of different
shapes for different purposes.
Each bridge will merit individual consideration
as to the most effective and economic arrangement and geometry of stiffeners. The Table
should, however, act ‘as a general guide.
Further detailed guidance is given in GN 2.03.
Table 1 Typical intermediate web stiffener configurations and detailing advice
Notes:
2.05/4
1. The 10 mm clearance is required to avoid a Class G fatigue detail See BS 5400-10, Table 17b.
2. The stiffener weld overlapping the longitudinal flange/web weld is now a widely accepted detail and is not
considered to carry a fatigue penalty.
3. Minimum sizes may need to be increased if there are gaps, for example between one end of a stiffener and
the flange where the stiffener is attached at both ends.
4. Welds should normally be continuous and ail round the stiffener.
©
2000 The Steel Construction Institute
Steel Bridge Group
Guidance Note
Welds - how t o specify
Scope
This note gives general advice on the way in
which t o specify welds in contract documents
and provides suggestions for further reading.
It is based on the assumption that all welds are
arc welds.
Terminology
There are a number of terms in this Note that
are used with specific meanings. Where these
terms are first used in the Note they are underlined. The standard giving the definitions is
BS 499-1 :1991, Welding terms and symbols:
Glossary for welding, brazing and thermal
cutting; it is worth studying the Standard to
avoid misunderstandings.
Other European and International standards use
somewhat different definitions and symbols, so
care is necessary when they are invoked or
cross referenced.
General
Guidance Notes GN 2.01 t o 2.05 give advice
about detailing of steelwork. They give guidance on h o w a bridge girder is assembled (i.e.
h o w it is joined together) and give advice on
where t o put splices in plates, stiffeners,
fittings and attachments. The pieces are
brought together t o be connected by joints.
These joints fall into seven defined types of
joints: butt, T, cruciform (a double T), lap, plug
(a form of lap), corner and &.
Most types of joint, including corner joint, can
be made b y one of t w o main types of weld:
butt or fillet (or combinations of both). Where
t w o plates lie flat together, with their edges
aligned, they can be joined b y a kind of butt,
defined as an edge weld. Plug welds or fusion
spot welds are used t o make some lap joints,
although not often in bridgework.
A butt weld is most commonly used in a butt
joint, t o join t w o plates which are co-planar, t o
make one plate the continuous and wholly
integral extension of the other. A fillet weld is
used around a lap joint and where one plate
meets another at or about a right angle (a
corner or T joint).
Sometimes, however,
because of design requirements, corner, T and
cruciform joints are required t o be made with
butt welds.
No.
2.07
The type of weld will depend most on its
function. In the simplest terms. a butt weld is
specified when the joint has t o transfer the
whole of the force in tension from one side of
the joint t o the other side, and/or when there
is a requirement for a joint of high fatigue
classification. Where there is no need t o
transfer the full force, the connection can be
made w i t h fillet welds, by butt welds of partial
penetration or with a compound of b u t t and
fillet welds. Such joints have an inherently
lower fatigue classification than properly
executed butt welds and may not be acceptable in certain locations on a structure. (See
clause 14.6.2 of BS 5400-3 for limitations on
the use of partial penetration welds.)
BS 499-1, Table 29, has sketches illustrating
the configuration adopted for making a number
of specific welded joints, 36 in all. It defines
the geometrical differences/ limitations applicable t o certain sub-groups of joints and includes
a remarks column covering the degree of
penetration required and the side from which
the joint is welded.
Execution - general
A weld is a particular way of making a joint; in
a number of instances there is no practical
alternative t o welding. A weld is, at best, a
necessary evil; it has costs in preparation, in
performance and in testing and inspection.
Even before commencing production work
there are costs in training and qualification of
welders, and in performance, testing and
certification of the particular weld procedures
t o be used (see GN 4.02).
Of the types of weld, butts are considerably
more expensive t o make than fillets. Butts
require particular edqe preparation of the fusion
faces and root face. These require accurate
forming, either b y flame cutting or b y machining. The t w o parts then have t o be brought
together within close tolerances of alignment,
in plane and in the gap between the root faces.
They have t o be held in that position until
sufficient weld is laid t o avoid relative movement. The volume of weld metal laid is significant, often requiring several runs. The welding
sequence is important: both t o avoid distortion
at the joint and t o achieve the specified mechanical properties of the weld metal and
adjacent parent metal. Finally, after the weld
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.07/1
Guidance Note
No.
2.07
is completed, the requirements for, and extent
of, inspection and testing are considerably
more for butts than fillets. They usually
involve the preparation and destructive testing
of test pieces (test coupons) for a representative sample of similar joints. In addition, the
requirements for inspection invariably include
sub-surface (sometimes called volumetric)
non-destructive examination of the weld metal
and the adjacent fusion zone and parent metal
t o confirm the integrity of the joint. This
inspection coverage for critical welds in tension
zones is often 1 0 0 %
Fillet welds, on the other hand, are usually
made at the junction of the flat surface of a
plate (which requires no preparation except
reasonable cleanliness) and the side(s) of
another flat plate (which only needs t o be
prepared square and straight). The requirement for reasonably close contact between the
parts t o be joined (the gap) is a positive advantage. The gap cannot be too small and sufficient contact is usually achievable by pressing
together the t w o elements, which is necessary
anyway t o locate the attachment in the correct
place. The control of the welding sequence is
usually less of a problem (most fillets can be
laid in one run) but it is sometimes necessary
t o balance the welding on each side of a joint
t o avoid distortion (rotation) of the attachment.
Finally, except for large fillet welds (more than
12 m m leq lenqth), the only inspection that can
be performed is visual (although this does
include checking the weld size) and
non-destructive testing for surface breaking
imperfections. Even with large fillet welds,
where sub-surface inspection is required, the
results need very careful evaluation by an
experienced welding engineer to avoid
unnecessary rejection because of spurious
responses from the root and toes.
Design
The application standard for the design of steel
bridges
(BS 5400-3)
lays
down
the
requirements for welded connections in Clause
14.6. This is used, among other things, t o
determine where types of weld may and may
not be used. It also covers such things as the
permissible strength/stressing and effective
l e n g t h s / t h r o a t thickness for welded
connections. It indicates certain requirements/
limitations on such things as continuity, or
2.07/2
otherwise, and deals with ends and returns of
welds.
A key requirement, in Sub-clause 14.6.1, is
that welds shall be detailed in accordance with
the appendices of BS 5 1 35. In the 1 9 8 4 issue
of BS 5135, Appendix A gives guidance on
design (of welds). Appendix B gives guidance
on butt welds and includes, among other
things, Table 15, which shows typical forms
of butt weld preparation (details) for joints in
materials other than structural hollow sections.
Appendix C deals with similar matters for
structural hollow sections. The sketches in
these Tables are generically similar to those in
BS 499-1, Figure 29, but have additional
dimensioned parameters and associated limits
or ranges thereof. The implication is that all
welds shown on contract documents should be
drawn and dimensioned in accordance with the
configurations in this Figure. This is not wholly
appropriate, however, for reasons given below.
Execution - specific
It is the designer's responsibility t o determine
the type of joint and the type of weld, the
extent of penetration if partial, and the size if
a fillet, at each location in the assembly.
However, the designer does not know how the
fabricator will chose t o perform the work: the
sequence of assembly, the order and
orientation of welding and the particular
welding process(es)/ welding sequence(s1 that
will be employed. These matters are within
the fabricator's competence and decision, to
enable him t o perform the work, achieving all
the specified requirements, in the most
economical manner of his choice. These
choices involve the exact shape of weld
preparation, root face, gap, included angle, and
are usually different for the same type of weld
when it is performed in a different way and/or
with a different process. Further, there are a
number of configurations in BS 51 35, Tables
15 and 16, that are, or can be considered,
alternatives.
The related standard for materials and
workmanship of steel bridges (BS 5400-6) also
invokes BS 5135, in its Clause 4.7.
In
sub-clause 4.7.1 it requires that "the general
procedures for shop and site welds, including
particulars of the preparation of fusion faces,
shall be submitted (by the Contractor/
© 2000 The Steel Construction Institute
Guidance Note
No. 2.07
Fabricator), in writing, in accordance w i t h the
requirements of BS 5 1 35, for the approval of
the Engineer before commencing fabrication".
However, Clause 3.1 (c) of BS 5 1 35: 1984,
dealing with information t o be supplied by the
purchaser (effectively, the designer), does
include "Locations, dimensions and details, i.e.
form of joint, angle between fusion faces, gaps
between parts, etc. of all welds".
It is suggested that these requirements are
somewhat contradictory and that welds should
be specified in contract documents as below.
Welds - how to specify
Designer: provide information on the location,
dimensions and types of all welded
connections.
For butt welds specify the
degree of penetration and, if partial, specify
the minimum design throat thickness. Specify
any particular requirements for the surface
finish of the weld profile. (The other matters
in BS 5 1 35: 1984, Clause 3.1 should already
be covered b y other parts of the model
contract documents but this should be
checked.)BS
5400,
Steel
Fabricator: provide inf o r m at io n on the
dimensions and details of the weld
preparations that are intended t o be used.
These should be in accordance with (i.e. within
the ranges given by) BS 5135. The Engineer
should check this. If not, the fabricator should
be able t o provide acceptable reasons t o the
Engineer for going outside the ranges, but may
still be required t o demonstrate acceptable
performance by carrying out appropriate tests.
See GN 4.02.
Both designer and fabricator: use an acceptable
system of welding terms and symbols, for
example those given in BS 499-1 and
BS EN 2 2 5 5 3 .
Other considerations
The foregoing is based on current UK practice.
European practice takes a less prescriptive line,
relying more on the acquired and formally
demonstrated experience and capability of the
welding establishment and, in particular, on the
competence of the welding supervisor t o direct
the workforce in the adoption of appropriate
processes and procedures.
This note deals with the specification of welds,
and does not cover rectifying defects that arise
during the execution of the welding. In effect,
the act of specification assumes that the weld
will be executed without significant departure
from the requirements w i t h regard t o size,
shape, penetration (or otherwise), surface
finish, reinforcement, etc. Clearly, however,
in a complete fabrication there will be
departures from intent during execution of the
work. BS 513 5 gives guidance o n acceptance
levels (maximum size of imperfections before
they are classed as defects) in Appendix H and
Tables 18 and 19; BS EN 2 5 8 1 7 gives
guidance on quality levels for imperfections.
Guidance Note 6.05 has been prepared t o
provide additional advice on the subject. The
designer should also consider the relevant parts
of the Model Appendix 18/1.
Reference standards
BS 499-1 : 1991, Welding terms and symbols.
Glossary for welding brazing and thermal
cutting.
BS 5 1 35: 1984, Specification for arc welding
of carbon and carbon manganese steels.
concrete
and
composite
bridges.
BS 5400-3: 2000, Code of Practice for
design of steel bridges.
BS 5400-6: 1999, Specification for
materials and workmanship, steel.
BS EN 1011: Welding, Recommendations for
welding of metallic materials.
BS EN 1 0 1 1-1 : 1998, General guidance for
arc welding.
(Other Parts t o be published.)
BS EN 12345: 1999, Welding: Multilingual
terms for welded joints with illustrations.
BS EN 2 5 8 1 7: 1992, Arc-welded joints in steel.
Guidance on quality levels for imperfections.
BS EN 22553: 1995, Welded, brazed and
soldered joints: Symbolic representation on
drawings.
Other references
For those w h o are interested in actually
designing welds, the following is a useful
simple guide:
A guide t o designing welds, Hicks J, Abington
Publishing, 1989.
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.0713
Steel Bridge Group
Guidance Note
2.08
Attachment of bearings
Scope
This Guidance Note gives information about
practical detailing for the attachment of bearings t o plate girder and box girder bridges. For
advice on bearing selection, see GN 3.03.
plate is therefore usually tapered (and in
bridges with integral crossheads is often
tapered in both directions). Tapered bearing
plates are also provided on railway bridges,
although the taper is normally small.
General
Bearings for steel girder bridges are of many
types, such as knuckle bearings, sliding bearings, elastomeric pot bearings or laminated
bearings. They can be proprietary products or
purpose-made (the latter are usually all-steel
bearings, such as knuckle bearings). The
design of bearings should be in accordance
with BS 5400-9, but the attachment of the
bearings t o the structure is n o t covered by that
Part. Unfortunately it is not specifically covered by BS 5400-3 either. This Note describes
attachment details that are commonly used,
and explains the design principles that are
normally applied.
There are t w o ways in which the bearings are
bolted t o the girders - using bolts through the
upper bearing element, the (tapered) bearing
plate and the bottom flange, or b y bolting
through the upper bearing element into tapped
holes. The design issues t o be considered for
these t w o methods are explained below.
Form of attachment
It is n o w a general requirement that bridge
bearings should be designed t o be replaceable.
This requirement may be more relevant t o
those bearings which have elastomeric elements that have a finite life, but it is a reasonable precaution in most cases. (Note that
BD 2 0 / 9 2 virtually makes it mandatory for
bearings t o be replaceable, because its universal replacement of 'should' in BS 5400-9.1 by
'shall' affects the recommendation in clause
5.4.)
Bolting through the girder flange
Simply bolting through all the elements is
perhaps the most immediately obvious means
t o attach the bearing. This makes an ordinary
bolted structural connection that can be designed t o BS 5400-3, provided that friction
grip bolts are used. With HSFG bolts in normal
clearance holes, there is no loss of effective
flange area in a compression flange (see
BS 5400-3, clauses 9.4.2.2 and 10.5.2).
Bolts are usually fixed with heads uppermost.
This facilitates the installation/removal of long
bolts and avoids tightening nuts onto nonsquare faces
The consequence of this requirement is almost
always that the bearings are bolted in place,
rather than welded.
Typically a four-bolt
arrangement is Provided on each of the upper
and lower elements. The bearing is bolted t o
the girder flange at the top and t o holding
down bolts set in reinforced concrete at the
bottom.
HSFG bolts can be used when the face under
the head is up t o about 2.5° from square t o
the bolt axis (i.e. about 4.5% gradient).
For several practical reasons, a 'bearing plate'
is usually Provided between the bottom flange
and the upper element of the bearing. In most
cases with parallel-flanged girders of highway
bridges the girder soffit is not horizontal (because the roadway is on a gradient or a vertical
curve), Yet the top surface o f t h e upper bearing
Plate almost always needs t o be set nominally
horizontal (to avoid horizontal components
arising from vertical reactions). The bearing
Difficulties arise, however, w i t h thick flanges
and moderately large gradient in elevation. For
example, a gradient of 2.5% and a flange
thickness of 60 m m reduce the effective
clearance of a vertical bolt in an inclined hole
by 1.5 mm. (During fabrication, it is only
feasible t o drill square t o the plate surface, not
at an angle.) This reduction in clearance would
make assembly difficult and it is better from
that point o f view t o use a slightly Oversize
Figure 1 Attachment by through bolting
SCI-P-185 Guidance notes on best practice in steel bridge construction
2.08/1
Guidance Note
2.08
hole. However, in that case the design resistance (to shear along the interface) is reduced
by virtue of being limited t o frictional resistance at ULS and the application of a reduction
factor k,. The effective area of the compression flange may need t o be reduced for the bolt
holes. Additionally, a top cover plate is,
strictly, required by clause 14.5.4.5 of
BS 5400-3, because oversize holes are acceptable only on inner plies, although thick washers
have been accepted as sufficient t o satisfy that
requirement.
Bolting using tapped holes
Bearing manufacturers consider that attachment using bolts into tapped holes is a satisfactory method of attaching bearings.
It
appears that grade 8.8 bolts are typically
suggested. The most common way of arranging this is t o tap the tapered bearing plate,
then weld it t o the underside of the girder
flange. The bolts are tightened but not
tensioned t o proof load (the bearing plate has
a much lower material strength than the bolt).
The problem with this detail is that it does not
comply strictly with BS 5400-3 on t w o counts:
non-preloaded bolts in clearance holes are not
permitted for structural connections; there are
no design rules (in Part 3) for the strength of
the connection (nor any specific requirement
for minimum engagement in the tapped hole).
Nevertheless, it is a commonly used detail, and
should be satisfactory provided that it is only
used where the horizontal restraint forces are
low (or zero, for unguided bearings) and the
following rules are observed:
Use grade 8.8 bolts. The size should be
sufficient for the design shear and at least
M20.
Use grade S355 steel for the bearing plate.
Ensure that a length of thread of at least
the bolt diameter is engaged in the tapped
hole
ciently to ensure that they do not work
loose- A torque value of at least half the
bedding torque for HSFG bolts (as specified
in the SHW, Series 1800 Table 2A) Should
be adequate provided that the threads are
undamaged and are cleaned by tapping Prior
t o installation. Alternatively use some form
of 'shake-proof washer'.
2.0812
Note that welding the plate t o the flange
creates a class G detail, which may govern
fatigue design at this position (especially at
intermediate supports or at fixed bearings of
railway bridges, where the shear forces are
high). In such cases bolting through the flange
may be preferable.
Figure 2 Attachment by bolts in tapped holes
(bearing plate is welded to flange)
ADDITIONAL
ATTACHMENTS
Whilst sliding bearings are best fixed with the
sliding surface facing downward so that extraneous material cannot fall onto the sliding
surface, they are often mounted 'inverted' t o
avoid eccentric loading t o the bearing stiffener.
in such cases skirts should be provided (to
prevent dirt, etc. reaching the upward facing
sliding surface). These can usually be attached
t o the upper part of the bearing, rather than
the girder flange. See Figure 3
Figure 3 Guided bearing (with skirt
turned back for inspection)
1. BD 20/92, Bridge bearings-use of BS5400:
Part 9: 1983, Department of Transport, 1992.
2. BS 5400-9-1 : 1983, Code of Practice for
design of bridge bearings.
3. Manual of Contract Documents for Highway
Works, Volume 1: Specification for Highway
Works, Series 1 800, Structural steelwork.
© 2000 The Steel Construction Institute
SECTION 3
MATERIALS AND PRODUCTS
3.01
Structural steels
3.02
Through thickness properties
3.03
Bridge bearings
3.04
Welding processes and consumables
3.05
Surface defects on steel materials
3.06
Internal defects in steel materials
3.07
Specifying steel material
Steel Bridge Group
Guidance Note
3.03
Bridge bearings
Scope
This Guidance Note provides information on
the types of bearings which are normally used
in steel bridges. The intention is t o provide
guidance as t o what type of bearings should be
used t o suit particular bridge forms. Fuller
descriptions of some bearing types, with
illustrations, are given in Lee (Ref 1) and the
Steel Designer's Manual (Ref 2).
In order t o select the most appropriate type of
bearings it is necessary t o consider the articulation of the bridge deck and its overall stability
under loading. Information on bridge articulation is given in GN 1.04. Information on the
attachment of bearings is given in GN 2.05.
General
In the majority of steel bridges it is convenient
t o use proprietary bearings. However, steel
fabricated bearings can often be economic,
especially in uplift situations, or where large
rotations occur, as in roll-on/roll-off link spans.
Where bearings have t o resist uplift, special
bearings normally have t o be designed and
manufactured. Hence uplift bearings are best
avoided, if necessary by modifying the design
concept. For lightweight structures such as
all-steel footbridges, consideration should be
given t o providing separate and robust holddown devices t o prevent dislodgement under
accidental impact or theft.
Whether bearings are proprietary or fabricated,
their design must conform t o BS 5400-9.1,
and their specification, manufacture and
installation must comply w i t h BS 5400-9.2.
accommodate rotation and, where required,
lateral movement in either longitudinal or
transverse directions, or in both directions.
Such bearings are particularly suitable for
continuous and curved viaducts.
For railway bridges or footbridges, fabricated
linear rocker bearings are often suitable at both
ends for simply supported spans up t o about
20 m. For rail bridges of span greater than
20 m, fabricated roller/rocker bearings can be
used at the free end. Linear rocker bearings are
of benefit in half through type rail bridges in
resisting transverse rotation t o achieve U-frame
rigidity if of plate girder type, or overall stability in the case of standard box girders with pin
connected floor.
For footbridges, elastomeric bearings are often
used. For short footbridge spans supported by
steel columns, bearings may be omitted altogether and replaced b y a direct bolted cap
plate connection, which allows thermal movements and articulation t o be accommodated by
flexure of the columns.
Some brief guidance on commonly used bearing types is given below.
Proprietary pot or disc bearings
These comprise a circular elastomeric disc
constrained b y a metal housing (forming a
cylinder and piston) which allows high pressures t o be used such that resistance t o articulation is negligible. Free bearings are achieved
by a PTFE/stainless steel interface, usually
arranged as shown in Figure 1.
top plate
sliding plate (stainless steel)
It is normal good practice for the designer to
make provision for the replacement of the
bearings during the lifetime of the bridge. This
usually means that, except for light
footbridges, provision for jacking should be
designed for and incorporated into the works.
Arrangements for fixing by bolting should be
such that the bolts can be removed without
needing t o raise the bridge girders.
Types of bearing
Table 1 shows the main types of bearing that
are used in steel bridges. The most frequently
used type of bearing for highway bridges is the
proprietary pot or disc type, which is able to
piston (with PTFE bonded on top)
elastomer
bottom plate
Figure 1 Free-sliding pot bearing
With this arrangement, when movement
occurs (due t o expansion/contraction) the
reaction becomes eccentric t o the superstructure above. This eccentricity can be avoided
by inverting the entire bearing assembly, but
the sliding surface is then vulnerable t o collecting debris, and should be shrouded by providing a skirt around the bearing (see further
comment in Figure 3 of GN 2.08).
SCI-P-185 Guidance notes on best practice in steel bridge construction
3.03/1
Guidance Note
3.03
The friction on the sliding surface depends on
the PTFE interface pressure and is typically
5%.
Pot bearings may be fixed, or guided, by
providing suitable lateral restraints between top
and bottom plates.
Proprietary elastomeric bearings
These may be of strip, rectangular pad or
laminated type (see Figure 2), the latter being
available typically up t o 1000 kN capacity.
Fabricated linear rocker bearings
These bearings provide a very economic
solution in that they can be supplied by the
steelwork fabricator. As well as economic
advantages that this gives, it may also give
more assurance that there is a good match
between hole positions in bearings, tapered
bearing plates and girder flanges, as well as
reducing the risk of delay in the procurement
process (see further comment below). See
Figure 4.
Figure 2 Laminated elastomeric bearing
Figure 4 Fabricated rocker bearing
For loads in excess of 1000 kN, the bearings
may become uneconomically large, such that
additional spreader plates and stiffening of the
steelwork become necessary. Thus elastomeric
bearings are rarely used for steel highway or
railway bridges. Design is governed by S L S
requirements, to control excessive distortion of
the material.
The rocker surface is normally radiused by
machine. The design capacity depends on the
radius, as well as on the material properties, as
given by the rules in BS 5400-9. 1 .
Movements and rotations are achieved by
deformation of the elastomeric material such
that moving parts are completely avoided. See
Figure 3.
Figure 3 Deformation of elastomeric bearings
I t is desirable t o retain bearings in position by
steel keep-strips attached t o the steelwork
above and the spreader p l a t e b e l o w . Movement is restricted to about 40 m m from the
mean position.
Elastomeric bearings are
unsuitable as fixed bearings, unless the forces
are small; some other form of restraint would
be needed to carry larger forces and this would
usually be expensive and might inhibit maintenance.
3.0312
A convenient method of attaching fabricated
bearings to the substructure in a way that
allows for large construction tolerances is t o
site weld the lower part of the bearing t o a
larger baseplate that is fixed t o the substructure before steelwork erection.
For skew plate girder bridges, linear bearings
should normally be aligned parallel t o the
supports so that articulation can occur normal
t o this direction. Otherwise uneven loading of
the bearing will occur, together with overstress
of the adjoining elements.
Roller/rocker bearings
These bearings are suitable for situations
where only longitudinal movement and articulation occur, and where transverse rotation is t o
be prevented. In order to reduce bearing
height, the upper and lower curved surfaces
can be radiused from different axes (see Figure
5), but this arrangement means that under
longitudinal movements, the superstructure will
tend t o lift, creating additional longitudinal
forces which must be designed for. This type
© 2000 The Steel Construction Institute
Guidance Note
3.03
of bearing is used at the free end of standard
box girder rail bridges for spans exceeding
2 0 m.
Figure 5 Fabricated rollerhocker bearing
Guides are provided t o ensure that the bearings
remain normal t o the direction of movement
and shear keys are provided t o transfer lateral
loads.
Proprietary roller bearings
Steel roller bearings usually comprise a single
cylindrical roller of high strength or case hardened steel to increase the load capacity and t o
minimise friction. This type is used where
friction is t o be minimised and is suitable on
leaf piers. It shares the displacement equally
between sub- and superstructure. The bearing
is normally enclosed within an oil-bath.
Guides are provided t o ensure that the bearings
remain normal t o the direction of movement.
Pin and swing link bearings
These bearings (also known as pendel bearings) are used where the amount of articulation
exceeds that of proprietary bearings as in rollon/roll-off link spans. They are also used for
footbridges in order to resist overturning and/or
theft.
-
Guide bearings
Occasionally, where lateral loads at a support
are large in comparison with the vertical loads,
separate guide bearings are used t o resist
lateral loads only. This situation can arise in
long viaducts where lateral forces are only
resisted at some supports, or in cable supported bridges. Proprietary sliding plate or pin
type guided bearings are available for small
loads, but guide bearings are usually purposedesigned.
Specification of bearings
The designer should prepare a bridge bearing
schedule according t o the format given in
BS 5400-9.1, Appendix A, Table 9. This
schedule is referred to by the supplier, and
gives a full statement of the performance
criteria including loadings, movements and
frictional resistance at all relevant limit states.
The design of the bridge including steelwork
stiffening, bearing plates and support on the
substructures, should be carried out by the
designer, assuming either one proprietary make
based on catalogue or other information from
an approved supplier, or that there are projectspecific bearings for which the full details are
given in the project documentation. The
drawings should then clearly permit the contractor t o offer different bearings, provided
that all design criteria can be met.
BS 5400-9.2 contains specification clauses
relating t o inspection and testing procedures,
including the provision of testing procedures
for proprietary bearings.
Procurement of proprietary bearings
Whilst the provision of a comprehensive bearing schedule gives a basis for competition
between bearing suppliers, it should be noted
that the time for gaining approval of selected
bearings is often very long. It is all too common for the bearing details to be outstanding
after the main body of the steelwork has been
fabricated - sometimes even after protective
treatment. It is essential t o progress bearing
procurement as quickly as possible, and t o
remember that it is false economy to save a
little on the cost of the bearings at the expense
of delay to the rest of the job.
Figure 6 Swing link expansion/uplift bearing
SCI-P-185 Guidance notes on best practice in steel bridge construction
3.03/3
Guidance Note
3.03
Robustness
Most bearings, purpose-fabricated or proprietary, are extremely robust and give little
trouble in service. The only problems that can
be said t o occur quite often are contamination
of, and hence damage to, sliding surfaces,
caused by leakage from the deck washing grit
onto the interfaces. This problem can be
avoided by careful detailing of bearing attachments and/or the provision of skirts (see
Figure 3 of GN 2.08 for an example of a skirt).
There have been some very occasional problems with elastomeric and hardened steel roller
or ball bearings caused by high frequency
vibration of the bridge structure. The rubber
degrades into dust; the steel surfaces develop
micro-cracking
Location plates for railway bridge bearings
It is normal practice for bearings for railway
bridges t o be positioned on location plates
bolted t o the abutment, particularly when
superstructures are reconstructed on existing
abutments during possessions. The lower
bearing plate is site welded to the location
plate. Location plates offer the following
advantages:
could be either proprietary or purpose-made;
provided that the steelwork is properly
protected against corrosion, they should be
maintenance-free.
In some types of jointless or semi-integral
bridge, sliding bearings might be used, with
longitudinal and lateral restraint being provided
by earth pressures on the endscreen walls
across the end of the superstructure.
References
1. Lee, DJ, Bridge bearings and expansion
joints, 2nd edition, E & FN Spon, 1994.
2. Matthews, SJ, Bearings and joints, Chapter
28 in the Steel Designers Manual, Blackwell
Scientific Publications (Fifth Edition), 1992.
3. BS 5400, Steel concrete and composite
bridges.
Part 9: Section 9.1 : 1983, Code of Practice
for design of bridge bearings.
Part 9: Section 9.2: 1983, Specification for
materials, manufacture and installation of
bridge bearings.
They provide tolerance on the final position
of the bearing relative t o the abutment.
Longer holding down bolts may be used t o
fix the location plates than could be used
for the narrower base plates (because of
the restricted clearance between the lower
bearing plate and the underside of the
bridge girder).
They achieve greater load distribution and
therefore reduce bearing stresses on the
abutment (this is particularly relevant on
brickwork abutments and may enable the
depth of sill beams to be reduced).
The bearing installation period required in a
possession can be minimised, because
location plates can be installed as soon as
the abutment is ready and before the bridge
superstructure is erected.
Bearings in integral bridges
In fully integral bridges, there is no translational
movement between superstructure and
substructure. In some types of integral bridge,
linear rocker bearings may be provided, to
allow relative rotation. Such rocker bearings
3.0314
© 2000 The Steel Construction Institute
Guidance Note
3.03
Table 1 Types of bearings
Type
Common
capacity
range (kN)
supply
Typical
friction
Pot or disc
500-30,000
Proprietary
0.05
Elastomeric
strip
200-1,000
Proprietary
Elastomeric
pad
10-500
Elastomeric
laminated
Cylindrical
roller
Widely used
4-10
kN/mm
Short span>
10m
Limited
translation and
rotation
Economic for short
spans
Proprietary
0.5-5.0
kN/mm
Short spans light loads
Limited
translation and
rotation
Useful light loads
100-1,000
Proprietary
0.5-5.9
kN/mm
Short spans
Heavy loads
Widely used
1,000-1,500
Proprietary
Nil lateral
translation or
rotation
Little used. Guides
essential
Cylindrical
knuckle
2,000-10,000
100-1000
0.01 (single Minimal
roller
friction
hardened)
0.25
Fixed bearings.
Rail bridges
High friction. Nil
lateral rotation
Large rotation
Fabricated
NA
Pinned
bearings. Base
of steel portal
Unsuitable
translation or
lateral rotation
Little used
Proprietary
0.005
Sliding guides
with large
translation
Small rotation
capacity
Suitable very short span
(say< 5 m) where
rotation negligible
0.05
> 20 m span
More expensive
than pot
Rotation capacity 0.05
Horizontal load
only
Carries no vertical
load
Used when guide
bearing essential, e.g.
end of long viaduct or
wide bridge
Fixed with
uplift
Nil translation or
lateral rotation
Useful footbridges for
security or uplift
Guided with
uplift
Nil lateral
translation or
rotation
Useful footbridges for
security or uplift
1 , 0 0 0 - 1 2 , 0 0 0 Proprietary
150-1,500
Proprietary
0.05
Pin
10-1,000
Fabricated
NA
Swing link
10-1,000
Fabricated
Control by
link length
Guide
General comments
Rotation capacity
0.01 radians
1 , 0 0 0 - 1 0 , 0 0 0 Fabricated
Spherical
sliding
Limitations
> 20 m span
Linear rocker
Plane sliding
Use
SCI-P-185 Guidance notes on best practice in steel bridge construction
3.03/5
Steel Bridge Group
Guidance Note
3.04
Welding processes and consumables
Scope
This Guidance Note is intended t o give the
reader who is not familiar with arc welding an
insight into the types of welding equipment
and consumables commonly used in steel
bridge fabrication.
Welding processes
Welding processes generally fall into one of
four principal groups:
manual metal arc welding
gas shielded metal arc welding
self shielding flux cored welding
submerged arc welding.
Within each group there are variations in
equipment and consumables, and there are
many individually identified sub-groups. The
most common processes are explained below.
Manual metal arc welding (MMA) process
This is a widely used and highly versatile
process. Although its use in shop fabrication
has diminished over recent years, it is still used
extensively on site.
The process is manually operated using individual flux-covered electrodes (often called rods)
that are inserted into an electrode holder held
by the welder. The process is sometimes
known as ’stick welding’. The resulting weld
is covered in slag that has to be removed
mechanically, usually with a chipping hammer.
Shop equipment usually operates on alternating
current (AC) via a mains transformer. The most
commonly used site equipment operates on
direct current (DC), powered by a diesel generator. However, AC diesel powered equipment
is also available.
There are advantages and disadvantages to
each process:
Virtually all electrodes will operate satisfactorily on DC, but special flux compositions
are required t o give a stable arc with AC.
AC equipment is generally simpler and more
robust than DC equipment.
AC equipment requires a greater open
circuit voltage, which may be a safety
problem outside the workshop.
The DC arc can be difficult to control,
because of magnetic effects. This is known
as arc-blow.
Electrodes commonly used are between
3.25 m m and 6 m m diameter and typically
4 5 0 m m long. The core wire is usually a low
alloy steel wire, alloy addition is usually via the
flux covering. Iron powder is often combined
in the flux t o improve deposition rates.
A wide variety of covered electrodes is available, from a considerable number of manufacturers, covering a wide range of welding
requirements.
‘Basic’ covered electrodes are normally used in
bridge construction. The main constituents of
the flux covering are calcium carbonate and
fluoride. Following proper drying and storage,
these electrodes are commonly known as
hydrogen controlled, and produce joints with
very good mechanical properties.
Basic electrodes must be stored with care to
ensure that they do not absorb moisture which
could generate hydrogen in the weld pool.
Storage in a temperature controlled oven is the
normal practice, with transfer t o heated quivers at the point of welding.
Consumable manufacturers now also offer
specially sealed packs that guarantee very low
levels of diffusible hydrogen if the electrodes
are used within the manufacturer‘s specified
period after first opening the sealed pack.
These packs are very useful if the amount of
welding t o be carried out is relatively small and
the setting up of a temperature controlled oven
would not be practical.
Deposition rates are relatively low with MMA,
usually in the range of 1 t o 1.5 kg/hr, as the
process is not continuous. For the typical heat
inputs and welding electrode diameters used in
bridge construction, the length of weld from
each electrode used in the downhand position
is in the order of 200 t o 3 0 0 mm, so there are
many stop-starts when using this process.
Gas shielded metal arc welding (MIG/MAG)
process
The process is mainly used in manual mode
and utilises a continuously fed consumable
wire from a wire feed unit to a welding gun
held by the operator. A shielding gas t o protect
the arc is also fed through the gun.
SCI-P-185 Guidance notes on best practice in steel bridge construction
3.04/1
Guidance Note
3.04
Continuously fed wire and shielding gas also
make this process well suited t o automated
welding.
The first shielding gas t o be widely used was
CO,.
Consequently the process was commonly known in its early years as CO, welding.
However, the use of CO, alone as the shielding
gas limits the process t o use at lower welding
currents.
A t low welding currents, the consumable wire
undergoes 'dip transfer' t o the welding pool.
The welding gun feeds consumable wire until
the wire tip touches the weld pool; short
circuiting then takes place, which melts the
end portion of the wire and breaks the arc
momentarily. The process repeats as the wire
is constantly fed. This mechanism is useful for
achieving good control in positional welding,
but can be prone t o lack of fusion in the flat
and horizontal-vertical positions.
Today the most commonly used shielding gas
is a mixture of argon and CO,. Using this
mixture dip transfer can still be achieved at
lower welding currents, but higher welding
currents and deposition rates can also be
achieved.
At high welding currents the transfer of the
consumable wire t o the weld pool is by spray
transfer where the consumable wire melts in
the welding arc and is transferred along the arc
to the weld pool.
After the advent of gas mixtures, the process
began to be referred t o as MIG (metal inert
gas) welding.
Later it was realised that the gases were doing
more than merely shielding the arc and that
changes to the gas mixture could affect weld
bead shape, deposition rates, fusion, and weld
metal properties. The process was then
termed MAG (metal active gas) welding.
Today it is possible that the reader may hear
the process referred t o by any of the three
(CO,, MIG or MAG), and to confuse the issue
still further the American term, gas metal arc
welding (GMAW), may also be used, but all
four processes are essentially the same.
3.0412
MAG is now the formally recognised terminology in the UK.
The gases used for bridgework are mostly
mixtures of argon and CO,. The proportion of
CO, varies between about 5% and 20%.
The MAG process is mainly used in the fabrication shop, although its use is growing on sites
in this country and it has been widely used on
sites in the USA for some years. The process
uses DC via a rectifier, and the electrode is
normally positive (DC ).
+
The most commonly used consumable wire
diameter is 1.2 mm. Wires are usually supplied
on 1 5 - 18 kg reels. Larger reels can be
obtained for use with robotic equipment.
Wires come in three main categories:
1j Solid wire, where the wire itself provides
the required alloys. Deposition rates in the
order of 3 kg/hr are typical.
2) Flux cored wires (flux cored arc welding,
FCAW), where the wire is hollow and filled
with flux. Alloying is derived from the flux
or wire, or both. These wires can give
deposition rates in the 3 t o 3.5 kg/hr range.
The type of flux can vary from rutile (titanium oxide), which gives the best positional
performance, t o basic flux (calcium carbonate) which gives better mechanical properties but is best suited to work in the flat
position.
3 ) Metal cored wires (metal cored arc welding,
MCAW), where the continuous wire is again
hollow but filled with iron powder. Alloying
elements are mainly contained in the continuous wire, but can also be present in the
core. Deposition rates of the order of
4 kg/hr can be achieved. MCAW can be
used for positional welding, but high skill
levels are required from the operator, and it
is therefore mostly used for work in the flat
position.
Development continues on suitable power
sources. There are at present t w o variants:
1 ) Synergic systems. These systems control
the melting process of the consumable
electronically, in order to mimic the characteristics and control of the dip transfer
mode but at higher welding currents and
© 2000 The Steel Construction Institute
Guidance Note
3.04
depositions. The operator has a single
control and all welding parameters are
automatically co-ordinated from that
control.
2 ) Pulsed arc. This system is intended t o
achieve the same end result as the synergic
system, but each welding parameter can be
controlled independently.
Self shielding flux cored process
This process (often known by the proprietary
name 'Innershield') is much the same as
FCAW, but relies on the flux contained within
the consumable wire alone t o generate the
shielding gas.
This process was very popular in the 1980s.
It gave good positional control and good
deposition rates compared with M M A welding.
The process develops relatively large amounts
of welding fumes, so much so that special
measures are needed for fume extraction. In
addition, the welding slag is classified as
contaminated waste and requires special
arrangements for disposal. As a result, use of
the process has diminished, but if the problems
are overcome in the future the process may
regain popularity.
Submerged arc welding (SAW) process
Submerged arc is a continuous welding process utilising solid wire electrodes where the
arc is submerged in, and is shielded by, a
granular flux.
The process is mainly used in automatic or
semi-automatic mode. Hand-held versions
have been developed, but are not widely used.
It is not suitable for positional welding; its use
is restricted t o the flat and horizontal-vertical
positions.
It is mainly used as a DC process and can be
operated as either electrode positive (DC + ) or
negative (DC-). The D C + mode is the most
commonly used and gives a high degree of
penetration. AC versions are available but are
not commonly used in bridge construction
Submerged arc welding is mostly used in the
fabrication shop, where a mains powered
rectifier unit provides the welding current.
It has been used on larger bridge sites for butt
welds in orthotropic decks. For this application
large diesel powered alternators are needed t o
feed the rectifier unit. The rectifiers usually
need some modification and weather proofing
t o function reliably and safely in the site environment. Large amounts of welding are necessary t o justify the use of SAW on site.
SAW equipment is commonly used mounted on
gantries, columns and booms, T&l plate girder
machines, deck panel assembly machines and
pipe welding machines.
High quality butt welds with excellent mechanical properties, and fillet welds with high penetration and good profile can be achieved with
this process.
Alloying generally comes from the solid wire
consumable, occasionally from the flux. Single
consumable wires of 4 mm diameter are the
most commonly used in bridge construction.
Deposition rates in the order of 6 kg/hr are
achieved in production.
T w i n wire systems w i t h the consumable wires
ed through a common welding tip are also
used and have the benefit of increasing deposition t o over double that of the single wire
system. Each wire is typically 2.4 m m diameter.
f
Tandem wire versions, where t w o wires feed
through separate welding tips, are also available and give deposition rates up t o 18 kg/hr.
A large number of flux/wire combinations are
available t o deal with a multitude of welding
situations and requirements.
Iron powder addition attachments which allow
deposition rates of up t o 20 kg/hr are also
available.
Most of the work and skill in using the SAW
process is in setting up the equipment. Therefore, even in the fabrication shop, it tends t o
be economic only for those joints requiring
large volumes of weld metal.
The process is quiet in operation, virtually no
fumes are created, and as the arc is totally
submerged in the flux no ultraviolet light is
emitted.
SCI-P-185 Guidance notes on best practice in steei bridge construction
3.04/3
Guidance Note
3.04
The operator does not require as much personal protective equipment (PPE) as with the
other processes.
Ceramic backings
Ceramic backing strips originated in the shipbuilding industry t o enable single sided butt
welds t o be made without the need t o fix and
leave in place steel backing strips. They have
been used extensively in bridge construction
over the last thirty years or so.
Ceramic strips of various shapes are available
t o fit in the roots of butt welds. The root run
is then deposited using the ceramic t o support
the weld pool. The ceramic can then be
chipped away after the weld has cooled.
High depositions are possible in the root run
and back gouging of the root can be minimised
or avoided.
Ceramic backings work well with the MAG
processes and MMA. With care, they can also
be used with the SAW process.
The ceramic backings are non-hygroscopic, so
the hydrogen performance of the welding
process is not impaired.
Unlike when using metallic backings, there is
no transfer of alloys t o the weld pool using
ceramic backings, there are no adverse effects
on the mechanical properties of the weld.
3.04/4
© 2000 The Steel Construction Institute
Steel Bridge Group
Guidance Note
3.06
Internal defects in steel materials
Scope
This Guidance Note describes the requirements
for limitations on internal imperfections in steel
plate materials. Mention is also made of
imperfections in rolled sections
The guidance relates t o steels manufactured in
accordance with current British and European
standards. Although the descriptions may also
assist in the assessment of older materials,
further guidance should be sought for such
purposes.
Terminology
Discontinuities within the body of the plate
were previously referred t o by names descriptive of the nature of the imperfection, for
example lamination, inclusion, inclusion cluster,
banding (of inclusions).
The present BSI
practice is t o use a generic term - internal
imperfections - and t o classify them by area,
length, breadth and population density
Note the distinction between an 'imperfection',
which indicates a feature that is less than
perfect but which does not prevent the component from being fit for purpose, and a 'defect',
which is a larger discontinuity that does impair
performance and may require rectification.
General
Steel plate is produced by a process of rolling
thick slabs of steel into approximately rectangular plates. During the process the material
is passed through rollers many times, each
pass resulting in a decrease in thickness and an
increase in area. Inevitably, the plate material
produced by this process is anisotropic. Any
discontinuities that may be present (for example resulting from bubbles or slag trapped in
the cooling slab) are elongated in the direction
of rolling and laminar in form.
Modern steels are generally very 'clean' and
the size of such imperfections is very small,
although they may occur in clusters.
In early steels, and particularly in wrought iron,
discontinuities were a significant proportion of
the plate area and the through-thickness performance of the plate could not be relied upon.
After rolling, plates are usually sheared or
flame-cut t o a rectangular size at the mills.
This may mask laminar imperfections at the
edge/ends, but any such imperfections would
be revealed when the fabricator trims the plate
Laminar imperfections can occur within the
body of a plate or at the edges. However,
even mid-plate laminations are likely t o appear
as edge laminations when the fabricator 'strips
out' a flange or web from the rolled product.
Laminations of significant size (area) would
impair the structural performance of welded
attachments t o the plate surface, and might
even give rise to local buckling failures in an
element under in-plane compression. Edge
laminations would reduce the integrity of
connections made at or near the edge of a
plate, for example in box girders or fabricated
channel sections. Edge defects would also
create a potential weakness against corrosion,
either as a result of entrapment of dirt or
moisture behind a coating, or by making a
discontinuity in the coating; any consequent
local corrosion would tend t o open the lamination and aggravate the problem.
Fortunately, laminar imperfections produced by
modern steelmaking processes are usually
modest in size and extent, and can be detected
by ultrasonic testing (and visually on clean
edges). The quality control procedures in
modern steel mills can generally be guaranteed
t o detect any gross defects and prevent them
from getting into the supply material. Problems usually only arise in material from older
(and often uncertified) mills, which may be
supplied through some stockholders.
Acceptance levels for internal imperfections
BS 5400-6 refers t o BS 5 9 9 6 for acceptance
levels for internal imperfections. Document
BS 5996: 1993 defines six basic acceptance
levels (designated B1 t o B6) for imperfections
within the 'material body', and three basic
levels for imperfections at the 'material edge'.
(There are t w o other levels relating t o
austenitic plate, but that material is not used
in bridgework.)
BS 5400-6: 1999 clause 3.1.4 specifies level
B4 and (incorrectly) 'level E' for certain areas
of plates. The designer has the facility t o
specify limits for other areas as considered
necessary. Note that the reference to level E
SCI-P-185 Guidance Notes on best practice in steel bridge construction
3.06/1
Guidance Note
3.06
mistakenly relates t o the earlier version of
BS 5996; the appropriate level according t o
the 1 9 9 3 standard should be level E l , which
corresponds closely t o the old 'E'.
However, BS 5 9 9 6 has recently been
withdrawn and has been replaced b y
The designation
BS EN 10160:1999.
equivalent t o B 4 is n o w 'S,', whilst equivalent
level for edge discontinuities is designated 'E,'
(rather than E l ).
Remedial/repair procedures
If a situation arises where unacceptable
defects are discovered at a stage in the work
when replacement is not practicable without
jeopardizing the completion programme, there
are various procedures that have been used
successfully in the past t o overcome the
problems.
Note, however, that it is important that any
repair conforms t o the same required standards
of workmanship as any other welding on the
fabrication. Special attention needs t o be
given t o the finishing and, where necessary,
surface dressing, t o ensure that the fatigue
class of the repair is no worse than that
required b y the designer at that particular
location. It is for this reason that no such
repairs should be carried out until the formal
approval of the Engineer/Designer is obtained:
as is required b y the Specification.
Repairs to edges
Shallow edge imperfections, particularly those
caused by the overlapping in rolling, can be
dealt w i t h b y grinding in from the edge t o
make a groove suitable for welding, and
employing an appropriate procedure t o weld
the edges of the plate together. Where the
edge imperfection is wider, usually part of an
area of de-lamination in the middle of a larger
plate which has been exposed by stripping into
narrower pieces, the edges can be bound
together as above and the wider area treated
as below.
Repairs in the middle of plates
In the very unlikely event of needing t o carry
out a repair t o the middle of a plate, the
affected area can be clamped together by use
of HSFG bolts or b y plug welding. If such a
repair is contemplated, the size and spacing of
3.06/2
any bolts or welds must be considered
carefully. Plug welding might be suggested for
better appearance, although it is very difficult
t o completely disguise a field of plug welds.
A better solution is t o cut out the affected area
and let in a new piece of plate, if this can be
done.
Testing
Any weld repairs should be thoroughly tested,
b y MPI for single run repairs and by MPI and
U T for deeper repairs.
Material with through thickness properties
A requirement for material with through
thickness properties is a different requirement
from that limiting the extent of internal
imperfections. However, specifying through
thickness properties in accordance w i t h
BS EN 1 0 1 6 4 invokes a requirement for
inspection for laminar imperfections of a high
sensitivity. Such inspections should identify
any significant imperfections.
See GN 3 . 0 2 for advice about the need for,
and the means of specifying, through thickness
properties.
Internal defects in rolled sections
Rolled sections are susceptible t o similar
problems but, once again, the quality controls
in modern manufacturing facilities should
prevent unacceptable material reaching the
market.
When dealing with older sections, or with
products from less well controlled mills (such
as might be supplied by a stockholder), there
are sometimes inclusions in the corner of an
angle or channel or at the web-to-flange
junction of a beam or Tee. Except in very thin
sections this would seldom be a structural or
durability problem, but might need t o be dealt
with where a gusset or cleat is attached near
the corner (angle/channel) or in the middle of
the flange (beam/Tee).
Safety note
The area of attachment of lifting lugs should
always be checked for sub-surface defects,
whether the material is plate or a section.
© 2000 The Steel Construction Institute
Guidance Note
3.06
References
1. BS 5400-6: 1999, Steel concrete and
composite bridges. Specification for
materials and workmanship, steel.
2. BS 5996: 1993, Acceptance levels for
internal imperfections in steel plate, strip
and wide flats, based on ultrasonic testing.
(Withdrawn).
3. BS EN 101 60: 1999, Ultrasonic testing of
steel flat plate product of thickness equal t o
or greater than 6 m m (reflection method).
4. BS EN 10164: 1993, Steel products with
improved deformation
properties
perpendicular t o the surface of the product
- technical delivery conditions.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
3.06/3
Steel Bridge Group
Specifying steel material
Guidance Note
3.07
(D07)
Scope
This Guidance Note describes the aspects that
need t o be specified t o ensure that the steel
material is of the strength and quality presumed when designing in accordance with the
design Standard (normally BS 5400-3).
GeneraI
BS 5400-3 presumes that steel materials will
be supplied in accordance BS 5400-6 Specification for materials and workmanship, steel.
In the latest edition (1999) BS 5400-6
Clause 3, Materials, states, in Subclause 3.1 .1,
that structural steels shall comply with the
requirements of a number of BS EN Standards.
These European standards have replaced the
earlier, all embracing, materials standard
BS 4360.
Included in the list of Standards are t w o that
cover modern fine grained steels; one of these
covers steels produced via a quenching and
tempering (Q&T) route. These t w o types of
steel come under a generic classification of
alloyed steels.
BS 5400-6 invokes BS 5 1 3 5 as its welding
standard. Clause 21.1 of BS 5 1 35:1984 sets
out the procedures for the avoidance of
hydrogen cracking, but also notes that in some
instances special measures need t o be invoked;
it directs the reader t o Appendix E. Of the
measures in E.3, items (cl), (d) and (e) do or
could apply t o alloy steels. Specialist advice
should therefore be taken before specifying the
use of alloy steels in case there are unexpected
and expensive consequences.
Where steels complying with the requirements
of other specifications are offered, BS 5400-6
requires, in Clause 3.1.2, that the performance
requirements listed in Table 1 shall comply
with those specified in the standards listed in
Clause 3.1.1. F o r most of the requirements
this is a sufficient control t o ensure that there
will be no problem in their use. However, with
respect t o weldability, the ‘indication‘ is only
the maximum carbon equivalent. As some of
the characteristics of a weldment, e.g. the
profile or surface finish, are sensitive t o the
actual chemical composition and/or the
particular concentration of certain elements,
substitution b y steels of a different
specification may have unexpected consequences.
Requirements t o be specified
There are four basic categories of requirements
that need t o be specified:
mechanical properties
metallurgical properties
dimensional properties
physical condition
These requirements are covered variously in
the Standard (Bs 5400-6), in the project
specification (including the SHW and the
Railtrack Model Clauses for specifying Civil
Engineering Works - Section 90) and in the
standards for materials, delivery conditions,
inspection, etc.
BS 5400-3 and BS 5400-6 both relate only t o
‘hot rolled‘ steel material. The use of cold
formed sections is not covered.
Mechanical properties
Mechanical properties are simply specified by
reference t o BS EN 10025, etc. (see GN3.01).
These standards include ’Options’ relating t o
the tests for properties, b u t none of these
Options need be invoked in normal situations.
For some applications, it is necessary t o ensure
that the material has adequate tensile strength
in the ‘through thickness‘ direction. GN 3.02
gives more advice about the subject and makes
reference t o specification in accordance w i t h
BS EN 10164.
Metallurgical properties
The t w o main properties that the SHW requires
t o be considered for the Specification are the
Carbon Equivalent Value (CEV) and the
chemical composition.
A suitably low CEV will enable the fabricator
t o reduce or remove the need for pre-heating,
a process that is very expensive and somewhat
difficult t o control. The EN Standards allow
the specification of a maximum CEV as an
‘Option’. For example, BS EN 1 0 0 2 5 allows
the option t o specify a value that is given in its
The SHW does suggest that
Table 4.
consideration should be given t o specifying a
maximum CEV for the thicker materials in
Grade S275 (over 5 0 mm), S355 (over
3 0 mm) and for all thicknesses of S 4 2 0 and
S460. Steel supplied b y the major European
steel producers easily meet the limits such as
SCI-P-185 Guidance Notes on best practice in steel bridge construction
3.0711
Guidance Note
3.07
in Table 4, b u t it is prudent t o include a
specific requirement, particularly where there
are small quantities of thick material - they
may be sourced from stockholders, and the
producer may not always be one of the major
suppliers.
The need t o specify the provision of the full
chemical composition should be restricted t o
the use of alloy steels, This is done b y
invoking Option 3 of Clause1 1.1 of BS EN
10113-1 and stating the number of samples
required and the elements t o be determined. It
should also be a condition of the use of steels
of specifications other than those listed in the
1999 edition of BS 5400-6.
Dimensional requirements
It should not be necessary t o include any
tolerances on plate thickness and other
dimensional tolerances in the project
specification. Such matters are covered b y the
various standards invoked elsewhere and by
clause 3.1.7 of BS 5400-6.
However, it is prudent t o be aware of these
tolerances when designing, particularly in
design of connections. The tolerances o n
plates do not usually give problems, although
extremely unfavourable combinations of
tolerances on thick plates can give steps at
bolted joint that are outside the allowable
assembly tolerances and which may need
additional packings.
Problems are more
common with rolled sections, both open and
closed; if rolled sections are t o be butted endto-end, for example, the only practical solution
is t o try t o secure all the material from a single
source and rolling.
Physical condition
Unless there are particular requirements for a
superior surface finish quality, there should be
no need t o make any additional reference t o
any other option for new material supplied by
a manufacturer. The relevant European
Standard is BS EN 10163, that is explicitly
referenced in the standards for technical
delivery conditions. If there is any doubt, the
suggested wording of Clause 3.105 of the
Model Appendix 18/1 can be used.
Sometimes, however, a proposal is made t o
use old material, from a stockholder or from
the fabricator's o w n stock, which has become
3.07/2
rusty. Commonly this arises from the
requirement t o provide a small amount of
material of uncommon thickness. Inclusion of
the wording suggested in Clause 3.107 of the
Model Appendix 18/1 invokes BS 7079 and
prevents the use of material with excessive
corrosion or pitting.
For information, BS 5400-6 requirements cover
all that is necessary and ensure that surface
defects, which arise in the rolling process, are
treated in accordance with the standards
designated in Clause 3.1.1 It goes further and
states that any repair b y welding of a surface
defect shall only be carried out with the
approval of the Engineer, using a procedure
complying with the requirements of BS 5 1 35.
However, as the current issue of the
Specification for Highway Works, Series 1800
requires such matters t o be covered b y the
Contract, b y putting t h e m in the Model
Appendix 18/1, it will be necessary t o consider
whether any areas, edges or ends of the
structure are particularly susceptible t o fatigue.
Such areas should be identified on the
drawings and instructions should be given t o
avoid any welding on them without specific
reference t o the Designer
GN 3.05 and GN 3.06 are relevant t o surface
and internal defects in steel materials, as
supplied, and should be consulted.
Inspection and testing
The latest edition of BS 5 4 0 0 - 6 covers all the
normal requirements for inspection and testing
of the finished product. This includes such
things as the marking for different grades,
testing for internal defects t o the appropriate
level in BS EN 10160 and the requirements for
the manufacturer's inspection certificate in
accordance with BS EN 1 0 2 0 4 .
When material is procured from a reputable
manufacturer, so long as it is ordered correctly,
all these matters will be available for checking.
If however, a request is made t o use material
from a stockholder or unknown or nonapproved source, it will be necessary t o require
inspection, testing and certification in
accordance w i t h the Standard. Certificates of
Conformity seldom contain much, if any, of the
detailed information that should be available.
© 2000 The Steel Construction Institute
Guidance Note
3.07
References
1. Specification for structural steelwork for
bridges: A model appendix 18/1, The Steel
Construction Institute, 1996.
2 . Manual of Contract Documents for
Highway Works, Volume 1 : Specification
for Highway Works, Series 1800,
Structural steelwork, The Stationery Office.
3. BS 5135: 1984, Specification for arc
welding of carbon and carbon manganese
steels.
4. BS 5400, Steel concrete and composite
bridges.
Part 3: 2000, Code of Practice for design
of steel bridges.
Part 6: 1999, Specification for materials
and workmanship, steel
5. BS 7079, Preparation of steel substrates
before application of paints and related
products. Introduction (Various Parts).
6. BS EN 10025: 1993, Hot rolled products
of non-alloy structural steels. Technical
delivery conditions.
7. BS EN 101 13, Hot rolled products in
weldable fine grain structural steels.
Part 1: 1993, General delivery conditions.
8. BS EN 1 0 160: 1999, Ultrasonic testing of
steel flat plate product of thickness equal
t o or greater than 6 mm (reflection
method).
9. BS EN 10163, Specification for delivery
requirements for surface conditions of hotrolled steel plates, wide flats and sections
(Parts 1 t o 3).
10. BS EN 1 0 164: 1993, Steel products with
improved deformation
properties
perpendicular t o the surface of the product
- technical delivery conditions.
1 1 .BS EN 10204: 1991, Metallic products
- types of inspection documents.
SCI-P-185 Guidance Notes on best practice in steel bridge construction
3.07/3
SECTION 4
CONTRACT DOCUMENTATION
4.01
Drawings
4.02
Weld procedure trials
4.03
Allowing for permanent deformations
Steel Bridge Group
Guidance Note
4.02
Weld procedure trials
Scope
This note is intended t o give the reader an
insight into the fundamental principles involved
in the formulation, testing and qualification of
weld procedures.
The requirements of BS 5135
BS 5 1 3 5 contains references t o the standards
covering the manufacture of various types of
welding consumables together w i t h the requirements for their satisfactory storage.
Weld procedure trials are expensive and time
consuming. Even on a small structure there
can be a large number of combinations of
variables. It is therefore usual t o make reference t o previously approved trials, and for
those trials t o have been conducted in such a
way that a range of production specifications
can be derived from them.
There are also general clauses setting basic
workmanship requirements for the execution
of welds. However the main body of the text
relates t o the development of weld procedures
for the avoidance of cracking.
Requirements
BS 5400-6, clause 4.7, requires that written
weld procedures in accordance with clause 2 2
of BS 5 1 3 5 be submitted for approval. Clause
4.7.3 calls for weld procedure trials in accordance with Bs 4870, if specified by the Engineer.
BS 5135 remains current, while BS 4870 has
been superseded by BS EN 288.
Terminology
What are referred t o as weld procedure trials
in BS 5 4 0 0 are n o w formally described as weld
procedure tests. The sequence of testing and
approval n o w described b y BS EN 288 is as
follows.
Having considered the technical and practical
requirements of a proposed welded joint or
range of welded joints, the welding engineer
will formulate a preliminary Weld Procedure
Specification (pWPS). This will be presented
in accordance with BS EN 288-1.
Based on the pWPS, weld procedure tests will
then be carried out in accordance with
BS EN 288-3. If the results are satisfactory,
a Welding Procedure Approval Record (WPAR)
will be prepared recording details of the weld
procedure and test results.
From each WPAR a number of Weld Procedure
Specifications (WPS) can be prepared, since a
single WPAR approves a range of variables
(see BS EN 288-3). All production work
should be in accordance with an appropriate
WPS.
Three categories of cracking are cited:
Hydrogen induced delayed cold cracking
Solidification cracking
Lamellar tearing
BS 5 1 3 5 was first published in 1974, and in
that version all of the above issues were
covered in terms of guidance notes given in the
appendices rather than expressed requirements.
The 1984 revision changed the situation, with
the provisions for the control of hydrogen
induced delayed cracking becoming expressed
requirements of the standard.
The measures for the avoidance of solidification cracking and lamellar tearing remained
only as recommendations in the appendices.
Regarding the mechanical properties of the
welded joint, BS 5 1 3 5 states that these are t o
be as required b y the application standard, but
gives no guidance o n h o w this might be
achieved.
In formulating a pWPS for a particular joint or
family of joints, the welding engineer must
therefore follow the expressed requirements of
BS 5 1 3 5 in respect of hydrogen induced
delayed cold cracking, but then rely on professional judgement regarding the attainment of
the necessary mechanical properties and the
avoidance of solidification cracking and lamellar
tearing.
In addition a competent welding engineer will
also invariably consider measures for distortion
limitation.
SCI-P-185 Guidance notes on best practice in steel bridge construction
4.0211
Guidance Note
4.02
Weld procedure tests
The weld procedure tests are carried out on
standardised test pieces as specified in
BS EN 288-3, clause 6.
The test pieces are required t o represent,
rather than totally replicate the joints t o be
welded in production. Their configurations are
standardised t o ensure that the necessary test
specimens can be obtained.
occasional need for a supplementary weld
procedure test.
If there is any doubt regarding the authenticity
of an existing WPAR, then fresh weld procedure tests should be carried out.
Weld Procedure Specifications
The review of relevant WPARs and the preparation of WPSs for production work is a complex matter and should always be carried out
by a qualified welding engineer.
A competent welding engineer will carefully
select the features of each weld procedure test
so that the resulting WPARs will approve the
full range of joints t o be encountered in the
production work, with the minimum number of
tests.
It must be remembered that a WPS is valid for
production welding only when all the parameters are within the limits of validity of the
various items of recorded data in the WPAR.
Weld procedure tests are usually witnessed by
an examiner/examining body who countersigns
the WPAR certificate.
Any changes in welding process, consumable
designation, and type of welding current,
always require retest.
Non destructive testing of the test piece t o
class B of BS EN 2581 7 is required, as well as
destructive testing.
Fabricator's equipment and operatives
It is prudent to ensure that the welding equipment used in production is properly calibrated
and maintained and appropriate t o the WPAR,
and t o check that welders' qualifications are
relevant t o the WPS and up to date.
The scope of destructive testing is set out in
clause 7 of BS EN 288-3. Figure 1 shows
examples of the test pieces required t o qualify
a butt weld.
Weld Procedure Approval Record (WPAR)
The WPAR is the formal record of a satisfactory weld procedure test. It records all the
parameters for the test and the test results. It
is certified by the examiner/ examining body
that witnessed the tests and carried out the
necessary mechanical tests and visual inspection.
Since the parameters for any test are variables,
the WPAR certifies a range of values for which
the test is considered t o be valid. (For details
about acceptable ranges, see BS EN 288-3,
clause 8.) This allows a number of Weld
Procedure Specifications t o be prepared from
a single test.
Specialist bridge fabricators usually hold a large
library of WPARs. Provided that the records are
properly witnessed by an examiner/ examining
body, they will form a satisfactory basis for the
development of the specific WPSs necessary
to carry out the production work, with only
4.02/2
The new EN standards require that, subject t o
specific conditions, the welders' approval
certificate be signed at 6-monthly intervals by
the employer's authorised signatory. Should
the employer wish t o prolong a welder's
approval beyond t w o years, records of volumetric inspection and tests must be maintained
and attached t o the welder's approval certificate for verification by the examiner/ examining body. This can be a problem for a fabricator undertaking a wide range of bridgework some procedures and equipment may not have
been used for some time.
Application tests/trials
The contractual requirements are set out
above. The Client cannot demand more unless
it is clearly specified and/or allowed for in the
Contract. Sometimes, however, while the
WPARs and the particular WPSs may seem t o
cover the essential features of all the joints in
the work, some applications may be constrained in some way, e.g. lack of clear access
or by one type of joint leading into another. In
such circumstances, it may be prudent t o
© 2000 The Steel Construction Institute
Guidance Note
4.02
specify (and pay for) extra Application
tests/trials. These trials are not covered by
BS 5 4 0 0 or any other Standard, but it has
been found worthwhile t o call for them in
particularly difficult circumstances, t o ensure
that the welds can be completed properly. The
trials are carried out on purpose-made,
non-standard test pieces that truly represent
the application. The testing of such pieces can
be limited t o visual and dimensional inspection,
though sectioning with the preparation of
macro specimens and hardness testing of the
weld and HAZ is also useful confirmation of
the quality of the weldment. These specimens
are also useful as comparative reference pieces
for the inspectorate during the performance of
the work.
References
1. BS 4870, Specification for approval testing
of welding procedures.
Part 1 :1981 (withdrawn), Fusion welding of
steel.
2. BS 5135: 1984, Specification for arc
welding of carbon and carbon manganese
steels.
3 . BS 5400, Steel concrete and composite
bridges.
Part 6 : 1999, Specification for materials
and workmanship, steel
4. BS EN 288, Specification and approval of
welding procedures for metallic materials
Part 1: 1992, general rules for fusion
welding.
Part 2: 1992, Welding procedures
specification for arc welding.
Part 3: 1992, Welding procedure tests for
the arc welding of steels.
5. BS EN 2 5 8 1 7: 1992, Arc-welded joints in
steel. Guidance on quality levels for
imperfections.
SCI-P-185 Guidance notes on best practice in steel bridge construction
4.0213
Guidance Note
4.02
4.0214
0 2000 The Steel Construction Institute
SECTION 5
FABRICATION
5.0 1
Weld preparation
5.02
Post-weld dressing
5.03
Fabrication tolerances
5.04
Plate bending
5.05
Marking of steelwork
5.06
Flame cutting of structural steels
5.07
Straightening and flattening
5.08
Hole sizes and positions for HSFG bolts
5.09
The prefabrication meeting
Steel Bridge Group
Guidance Note
5.03
Fabrication tolerances
Scope
This Guidance Note presents a brief consideration of what can be expected in terms of the
accuracy of a structure's fully assembled
length, width and vertical profile.
Background
Dimensional tolerances are set out in
BS 5400-6, Table 5. These tolerances are
acceptable imperfections within components
where the strength of the component is
affected by imperfection; the magnitudes of
the tolerances relate t o assumptions of
imperfection made in the design code,
BS 5400-3.
No general dimensional tolerances are specified
for either components or fully assembled
structures.
Additionally, clause 6.3.1 of BS 5400-6 sets
out the general obligation that "bridge
steelwork shall be erected, adjusted and
completed in the required positions t o the lines
and levels specified by the Engineer ...".
However, it should not be taken from this
statement that bridge steelwork can be
assembled t o precisely the given dimensions,
with zero tolerance.
There are other standards in related industry
sectors giving specific dimensional tolerances
for individual components, and these can be
useful in assessing the general standard of
fabrication in individual components, but
should not be used as the basis for establishing
overall structure tolerances.
At a cost, and given the time, very high levels
of dimensional accuracy are possible, but the
main issue is what can be justified for the
particular structure.
Types of dimensional imperfection
As with all manufactured items, fabricated
components contain both random and
systematic dimensional errors.
On larger structures, where many girders may
be joined end-to-end, there will be a general
tendency for the random errors to be self
compensating: the statistical probability of all
such errors accumulating in one sense will be
very small. Systematic errors on the other
hand, will generally accumulate.
The relationship between the magnitudes of
random and systematic errors varies among
fabricators, depending on their methods and
their equipment, and will also vary with
structural form.
In bridge construction, length is one of the
most important dimensions that needs t o be
controlled. The overall dimensions of smaller
structures, with few fabricated components in
their length, will be influenced by both random
and systematic errors; longer structures, with
many components, will be influenced mainly by
systematic errors.
Generally it can be expected that the ratio of
total error to total dimension can be expected
t o be better on larger structures where there
are more components and the random
elements have a chance t o cancel one another
out, than on smaller structures with fewer
elements where there is less chance of the
random errors self compensating.
Component tolerances
BS 5400-6
As mentioned earlier, BS 5400-6 gives no
guidance on general dimensional tolerance (i.e
other than on dimensions affecting strength),
and for many years the question of what
constitutes an acceptable standard of
fabrication for bridgework has been open for
debate.
National Structural Steelwork Specification
The NSSS, first published by the BCSA in
1989, gave some assistance, but it was
drafted for the building construction industry,
not with bridge structures in mind. In the
NSSS, the length tolerance for individual
members is set at + 3 mm, which reflects the
requirements of girders of generally shorter
lengths than are used in bridge construction.
The NSSS gives tolerances for such matters
as: girder depth on centreline, flange width;
squareness of flange to web; eccentricity from
intended position of web from edge of flange;
out of flatness of flange (transversely); position
of load-bearing components such as bearing
stiffeners; position of holes. The values in the
SCI-P-185 Guidance notes on best practice in steel bridge construction
5.0311
Guidance Note
5.03
NSSS could be adopted for general tolerances
on structural bridge components.
themselves, and these should be provided with
some adjustment facility.
BS EN IS0 13920
Standard BS EN I S 0 1 3 9 2 0 gives general
tolerances for welded constructions. This new
standard has been drafted with the intention of
setting tolerance classifications that will cover
the whole steelwork fabrication industry. The
particular class t o be adopted would depend on
the end use.
If structural stiffeners are intended t o serve a
dual purpose (i.e. for attachment of fittings as
well as for structural purposes), this should be
clearly specified, so that any implications on
tolerance can be identified.
It must be stressed that these tolerances
should only be used t o assess individual
components. It should not be argued that they
can be expected t o accumulate arithmetically
when a significant number of such components
are connected.
Tolerances on length are set out in four
classes: Class A of this standard has been
adopted for bridgework by some European
countries.
However, whilst the Class A
tolerances (the highest class) are satisfactory
for bridgework dimensions over 4 m, those for
dimensions under 4 m are unnecessarily tight.
Tolerance classifications B, C and D are
unsuitable for bridge construction.
BS EN I S 0 1 3 9 2 0 also gives classifications for
straightness, flatness and parallelism, but these
requirements are already covered satisfactorily
and comprehensively by BS 5400-6.
Attachment of other prefabricated elements
It is necessary t o give consideration to the
accuracy of location of fittings required t o
support and/or avoid other prefabricated
elements, including non-structural items such
as, for example, parapets and fascia panels.
BS 5400-6 only addresses this matter
indirectly, in clause 4.2.1 General. The NSSS
gives a general tolerance for the location of
structural fittings as & 3 mm, which is related
t o their required location shown as a simple
dimension on the design and/or fabrication
drawings. Whilst this degree of accuracy is
reasonably easy t o achieve within a single
fabricated element, the consequences Of
accumulated errors over the joints in a series
of consecutive elements should be taken into
account in the detailing of the fixings
5.03/2
Adjustment of errors
BS 5400-6, clause 6.3.1 calls for adiustment
to achieve the required lines and levels. The
t w o aspects that most commonly require
adjustment are level (usually of the top flange)
and overall length relative t o the substructure.
On long-span bridges, adjustment may be
achieved by match marking and trimming joints
in rolling assemblies at ground level, taking
account of as-built surveys of the substructure.
This approach is slow and expensive, but is
essential for long structures (over about
2 5 0 m).
In viaduct work, adjustment points are
normally pre-planned, and the main girder joints
at those positions are not completed until
dimensions are available from site indicating
how the structure is matching the local bearing
locations.
Such adjustments are assessed in relation to
the substructure grid lines, which are
themselves subject t o tolerance, rather than
being absolute overall dimensions. Given a
good standard of fabrication and substructure
set-out, such adjustment points would typically
be at 1 5 0 m intervals.
Between adjustment points, errors accumulate.
The amount of error that can be safely
accommodated at each support position should
consider the following:
The allowable eccentricity in the support,
which is a matter of design.
The available spare movement capacity in
sliding bearings, which is a matter of
bearing selection and design.
Any implications regarding expansion joint
movements.
The flexing of tall slender piers when fixed
bearings are used
© 2000 The Steel Construction Institute
Guidance Note
5.03
The selection of adjustment points in long
structures s h o u l d t h e r e f o r e b e a m a t t e r o f
discussion a n d a g r e e m e n t b e t w e e n t h e
designer, t h e m a i n c o n t r a c t o r a n d t h e
fabricator, and should take place as early as
possible in the project.
value. The real issues are: what relative errors
can be safely accommodated between the
steelwork, the bearing and the support; and
can modes of adjustment of the appropriate
magnitude be established to accommodate
such differentials?
The assessment of misalignment at each
support as construction proceeds needs careful
observations, and should take into account
average steel temperature together with any
rotations and translations that will occur under
application of full dead load. The problems in
assessing temperature effects are discussed in
GN 7.02
This is reflected in the commentary to clause
The parties to a project are best
6.302.
advised to discuss these issues openly and
thoroughly at an early stage rather than seek
to set unrealistic and perhaps unnecessary
tolerances upon each other.
As bearing locations may have to be adjusted,
it is recommended that only fixed bearings be
fully grouted prior to erection. Pockets for
bearing fixing dowels should be detailed to
allow such adjustment of bearing base plates.
Temporary bearing packing systems should be
planned carefully in terms of capacity and
disposition in order to safely support the steel
dead weight in all erection conditions
Tolerances on length
Experienced bridge fabricators are generally
conscious of how tolerances within their
steelwork combine and have their own internal
rules based on experience of where
adjustments may be necessary. Figure 1 gives
an indication of the misalignments between
steelwork and substructure that can be
expected, given a good standard of fabrication
and substructure set-out but without special
adjustments.
Variation on Ls : ± (10 mm + Ls/10 000)
(LS is measured from the substructure to a point on
the superstructure.)
Structure width
Errors in width will tend to be higher,
relatively, than those in length.
This is
because there is usually a high number of
joints (per unit width) and any dimensional
errors tend to be dominated by systematic
effects.
Bridges skewed in plan will tend to have
greater relative error in width than those that
are square in plan.
A general tolerance of width/1000 should be
achievable.
Vertical profile of steelwork
The weld shrinkages at the top and bottom
flanges of bridge girders are often different,
because of differences in weld size and flange
cross section, and due to the sequence of
welding.
Such differentials give rise to a
vertical curvature in the girder, and this may
well be in the opposite direction to that of the
vertical profile a n d t h e a l l o w a n c e f o r
permanent deformation under dead load.
Fabricators use empirical rules to adjust the
shape of the web to allow for such effects as
shrinkage. This is not a precise science, and
the control of the profile, particularly on
slender girders, i s a m a t t e r t h a t r e q u i r e s
experienced judgement by the fabricator.
Figure 1 Expected deviations on length
Model Appendix 18/1 clause 6.302 iii)
suggests that the tolerance for plan position of
steelwork be set at ±5mm at datum
temperature. Establishing the absolute error in
the erected steelwork at datum temperature is
difficult and time consuming, and of limited
To complicate the issue further, all ‘as-welded’
fabrications have locations adjacent to welds
where residual stresses are at or about the
yield point for the material (the forces are in
w i t h o t h e r internal forces).
equilibrium
Externally applied vibrations or loads can give
rise to stress relief at these locations, and to
SCI-P-185 Guidance notes on best practice in steel bridge construction
5.03/3
Guidance Note
5.03
a resultant overall change in shape of the
fabrication as the equilibrium in internal forces
changes. This effect is not significant in the
majority of bridge girders, but can be an issue
in structures where the span t o depth ratio is
Girders of such
around 30 or more.
proportions have been known t o alter in their
profile during transportation.
For the above reasons, relative errors in profile
tend t o be greater in shorter spans. In longer
spans, the girders tend to be inherently less
slender, but also, since more individual girder
pieces tend t o make up the span and provide
an opportunity for vertical rotational
adjustment at each joint, considerably lower
relative errors can be achieved in the overall
profile of the span.
A tolerance of + span/1000 (on midspan level,
relative t o the level at the supports, reducing
proportionately as the distance t o the support
reduces) can be achieved; a tolerance of
3 5 mm can be achieved o n spans exceeding
3 5 m. This is the tolerance given in the Model
Appendix 18/1, but the designer can specify
lesser values in situations which require tighter
tolerances.
Designers should, however,
recognise what can be readily achieved by the
fabricator/erector and not specify tighter
tolerance except where strictly necessary.
’
In specifying the required tolerance for any
particular project, the designer should consider
carefully what the effects that deviations up t o
that tolerance value would have on vertical
profile, drainage, dead weight, surfacing
thickness, etc.
Note, however, that the above discussion
about tolerance o n level relates t o the bridge
as a whole, n o t t o each individual girder
separately. Where there would be a lack of
planarity at top flange level of a composite
bridge, the need t o maintain a minimum
structural thickness of the slab will usually
mean that the slab will be cast with its nominal
depth over the highest girder, and w i t h a
greater depth over the lower girders. In the
Model Appendix 18/1, clause 6.302 ii)
specifies a tolerance of 20 m m on the level of
one main girder relative t o another, adjacent,
main girder; this tolerance is appropriate for a
slab of about 2 5 0 m m thickness. If any
5.03/4
greater tolerance were t o be accepted, the
design should be checked for the consequent
increase in dead load due t o the thicker slab.
Profile of road surface
The profile of the finished road surface of a
highway bridge depends on three elements:
the profile of the steelwork
any variation in the thickness (and weight)
of the deck slab
any variation in thickness of the roadway
surfacing
As mentioned above, there may be some scope
for adjustment of steelwork during its
construction if it is expected that the deviation
from the required geometry will be too great.
If the steelwork is not adjusted, there is scope
for adjustment of profile by ’regulation’ of the
deck slab and/or the surface thicknesses. In
these cases the thickness is varied (usually
increased at ‘low spots’) so that a better
finished profile is achieved.
However,
additional thickness means additional dead load
and this will increase dead load bending
moments. Additional thickness and load in
midspan regions has a much greater influence
on moments than does an increase local t o the
supports. It may therefore be prudent t o ‘overcamber’ the steelwork in some cases t o guard
against variations (from the intended profile) on
the l o w side. Such addition would be made t o
the specified ‘allowance for permanent
deformation’ (see GN 4.03). If such a strategy
is adopted, it should be made clear t o all
parties.
In railway bridges, deviations between
intended and actual profile of the structure will
be accommodated b y variations in ballast
thickness but, again, the consequences o n
dead load and on the level of the track need t o
be considered.
Verticality o f girders a t supports
The verticality of girders at supports is usually
set by bracing or cross head systems
connected w i t h HSFG bolts. Given a good
standard of fabrication there is usually
sufficient clearance in the bolt holes t o allow
the appropriate degree of adjustment t o
achieve the tolerance of D / 3 0 0 without
resorting t o reaming.
© 2000 The Steel Construction Institute
Guidance Note
5.03
It is emphasised that this check applies t o the
web itself, and should not be translated to, or
derived from, the horizontality or otherwise of
the flange plates or bearings.
If this tolerance is not achieved and subsequent adjustment of the bracing system is not
readily achievable, a check will be necessary
t o establish whether the bracing system is
capable of sustaining the resulting additional
horizontal forces. This situation may well be
experienced at the abutments of skewed
composite bridges. See comment on girder
t w i s t in GN 1.02.
References
1. BS 5400-6:1999, Steel concrete and composite bridges. Specification for materials
and workmanship, steel
2. BS EN I S 0 13920:1997, Welding. General
tolerances for welded constructions. Dimensions for lengths and angles. Shape and
position.
3. Specification for structural steelwork for
bridges: A model appendix 1811, The Steel
Construction Institute, 1996.
SCI-P-185 Guidance notes on best practice in steel bridge construction
I
5.0315
Steel Bridge Group
Guidance Note
5.09
The pre-fabrication meeting
Scope
This Guidance Note suggests an outline agenda
for the prefabrication meeting between the
Contractor, the Engineer's representative and
the Inspector's representative.
The use of the 'Engineer' relates t o the
traditional
contractual
arrangements.
Appropriate changes of relationships can be
substituted for other forms of contract.
The 'Inspector' is taken t o be the appointed
inspection authority.
The inspector who
actually carries out the work is therefore
referred t o as the Inspector's representative.
The meeting
A t the commencement of a contract that
involves the fabrication of bridge steelwork,
the parties involved need t o meet at an early
stage t o discuss the clarity of the contract
provisions with respect t o all the relevant
technical issues. The principal parties are the
designer, the fabricator and the inspector;
where the fabrication is sub-contracted, a
representative of the main contractor should
attend.
2.0
Fabrication programme
3.0
Specifications for materials and
workmanship
Application code and British Standards
Use of other specifications and,
certificates of conformity (where
permitted)
3.1
3.2
4.0
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
Sub-contracting
Flame cutting
Fabrication and welding
Bending
Machining
Non-Destructive testing
Destructive testing
Protective treatment
Site erection, welding, non-destructive
testing and protective treatment
5.0
5.1
Materials supply
Material specifications/codes and steel
grades
5.2 Mills inspection
5.3 Supply condition
5.4 Supply by stockists
5.5 Certificates of test
The arrangements for carrying out the work
need t o be explained and recorded; access for
inspection needs t o be agreed; procedures for
dealing with queries that arise need t o be
established.
6.0
6.1
An agenda that was drawn up by Messrs
Sandberg for such initial meetings has been
adopted widely within the steel bridge sector.
A copy of that agenda is presented below,
with a few additional points t o cover increasing
use of quality assurance schemes.
6.4
Materials at Contractor's Works
Contractor's quality control system for
receipt
Contractor's traceability system
Certificates of test for Inspector's
representative's review
Lamination checks
7.0
Drawings
8.0
8.1
8.2
8.3
8.4
Preparation of materials
Flame cutting
Preparation of edges, ends and
surfaces
Fitted stiffeners
Holes for bolts
9.0
9.1
9.2
9.3
9.4
9.5
9.6
9.7
Welding
Welding procedure specifications
Welding procedure approvals
Welder approvals
Welding preparations
Temporary attachments
Welding distortion
Welding stud shear connectors
Careful consideration will have t o be given to
contractual arrangements encompassing quality
assurance and other forms of self certification.
This will need t o cover the intervention by the
designer at clearly defined review and 'hold'
stages, where necessary.
Agenda
© 2000 The Steel Construction Institute
6.2
6.3
5.09/1
Guidance Note
5.09
10.0
Bending
11 .O
Straightening and flattening
12.0
12.1
Production test plates
Location and numbers of production
test plates
Testing of production test plates
19.0
19.1
19.2
12.2
13.0
13.1
13.2
13.3
13.4
13.5
13.6
Non-destructive testing of welding
Methods of and areas requiring NDT
for butt and fillet welds
Acceptance levels for NDT
NDT equipment calibration
NDT operator qualifications
NDT procedures
NDT reports
14.0
14.1
14.2
Testing of stud shear connectors
Ring testing
Bend testing
15.0
15.1
15.2
Tolerances
Tolerance requirements
Temporary erection at contractor's
works
Contractor's method for recording
tolerance checks
Inspector checks
15.3
15.4
16.0
16.1
16.2
17.0
19.3
19.4
19.5
19.6
19.7
20.0
General points
Facilities for Inspector's
representative
Contractor's QC documents review
by Inspector
Contractor's quality plan
Inspector's quality plan
Inspector's unconfirmed reports
Inspector's representative's contact
with Engineer
Procedure(s1 for dealing with
imperfections, defects and other nonconformities
Any special requirements or
considerations
Bolts, nuts and washers
High strength friction grip bolts, nuts
and washers
Methods of tightening and pre-load
control for HSFG bolts
Handling and stacking
18.0
18.1
18.2
Protective treatment
Protective treatment system
Paint manufacturer and paint data
sheets
18.3 Paint storage
18.4 Testing of paints
18.5 Contractor's quality control system
for painting
18.6 Painting environmental conditions
18.7 Painting procedure trial
18.8 Blast cleaning reference sample panel
18.9 Metal spray reference sample panel
18.10 Damage t o protective treatment
18.1 I HSFG joint faying surfaces
18.12 Inspector's representative's
involvement in protective treatment
5.0912
…©2000 the Steel Construction Institute
SECTION 6
INSPECTION AND TESTING
6.01
Weld quality and inspection
6.02
Surface inspection of welds
6.03
Sub-surface inspection of welds
6.04
Hydrogen/HAZ cracking and segregation cracking in
welds
6.05
Weld defect acceptance levels
SECTION 7
ERECTION AND IN SITU CONSTRUCTION WORK
7.02
Temperature effects during construction
7.03
Verticality of webs at supports
7.04
Trial erection and temporary erection
7.05
Installation of HSFG bolts
7.06
Transport of steelwork by road
7.08
Method statements
Steel Bridge Group
Guidance Note
7.02
Temperature effects during construction
Scope
This Guidance Note provides some general
information about the effects of temperature
variation during bridge construction.
Temperature variation causes the bridge to
change length and, t o some extent, t o change
shape (in profile and plan). This may require
action to allow for the changes and for the
differences between actuality and design
assumptions; some general advice is given.
Design
The formal rules for establishing effective
bridge temperatures for the design of any
particular bridge are contained in BS 5400-2,
clause 5.4. These rules provide appropriate
values for the required range of movement of
such items as the bearings and expansion
joints. A more detailed background is given in
a number of TRRL publications (ref 1).
However, the presumption in design, and in the
calculation of the range of movement, is that
the bridge will be constructed t o the specified
length at 'datum temperature'.
Further,
movement ranges at expansion joints and
bearings generally presume that the
substructure and superstructure dimensions
will be the same at this temperature.
The question thus arises as t o what measures
need t o be taken during construction, at the
key stages when dimensions are fixed, to
establish the actual temperature of the bridge.
When the temperature is not uniform,
questions arise not only over length but over
other dimensions (transverse, rotational, etc)
when settings have t o be fixed.
Considerations
The principal considerations in establishing
effective temperature are:
of that face to the incident energy. This
means that low early or late sun or even
low midday sun in the winter, will have a
significant effect on the vertical faces and
high summer or midday sun will have an
effect on the horizontal surfaces of a
structure.
The effects of solar radiation are likely t o be
experienced disproportionately by different
parts of the structure, leading t o differential
temperatures in the elements, with resulting
distortion.
The nature of the distortion is extremely
difficult to predict. Hot sun shining on the
deck of a deep girder bridge will cause the
whole structure t o expand and the girder t o
tend t o hog between supports, but the
effect will be unequal on a long main span
compared t o a short side span. This may
actually cause the bearings t o move in
entirely the opposite direction t o that
expected in other conditions.
Notwithstanding the above, open steelwork,
such as a truss, responds very quickly
(within an hour) t o all changes in ambient
temperature, whether exposed t o sunlight
or not. Closed box sections tend t o have a
significant lag in response (3 t o 5 hours).
0bservations
A number of observations made over the years
can be useful in determining whether there is
likely t o be a difficulty with a particular
structure:
On very long structures, there can be
significantly different effects at different
cross sections, especially if over different
topography.
Narrow decks are more susceptible t o
lateral distortion than wide decks.
Solar radiation. Of the effects causing
changes in effective temperature, the total
solar radiation is the most powerful and
rapid. The effect is so strong that it does
occur even through ordinary cloud cover.
Deep girders behave more unpredictably
than shallow girders.
Box girders are more problematic than plate
girders, and they can bend considerably in
both the vertical and horizontal directions.
The effects are quicker to be felt and
quicker to dissipate in steel elements than
in concrete. This effect is exaggerated if
the steel is painted a dark colour.
Wide concrete decks which protect the
steelwork from direct sunlight reduce the
major effects.
Structures with significant heat-sinks
respond more slowly than plain steel
skeletons.
Concrete decks have a
7.02/1
SCI-P-185 Guidance notes on best practice in steel bridge construction
Clearly, the response of faces exposed to
solar radiation is related to the orientation
Guidance Note
7.02
significant damping effect and the air inside
box girders tends t o slow and reduce the
effects.
Thin plated structures react more quickly to
radiant heat and reach higher temperatures.
A concrete deck surfaced with 'black-top'
will react more than an unfinished concrete
surface.
An unsurfaced steel deck will respond more
quickly t o radiant heat and can reach higher
temperatures than even a black-top
surfaced concrete deck
As the effects vary during the day, it is
particularly important t o be aware of them
during the making of the site joints. Bolted
joints could be locked Up with unintended
distortion built into the joint; incomplete
welded joints could become over stressed
and fail during the night following a hot,
sunny day.
All these effects are likely to be more
troublesome
with partly completed
structures, i.e. during the erection process.
Towers and piers react t o diurnal variations
in the position of the sun.
Assessing effective temperature for Setting out
purposes
It is extremely difficult t o make use of the
design rules during the construction stage of
any particular bridge project: in particular, with
respect t o BS 5400: Part 2, Clause 5.4.4. This
is because the t w o major influences (direct
solar radiation and the subsequent dissipation
of the stored thermal energy from the structure
by radiation, conduction and convection) are
transient and variable. However, this gives a
clue as t o how t o deal with the problem, i.e.
by reducing the effect of these influences to a
minimum whenever one is faced with the need
to know the effective temperature of the
bridge or parts thereof.
The most reliable information is acquired by
avoiding the periods of day when solar
radiation is present and also avoiding
circumstances where the rate of change of
temperature is significant.
three o'clock in the morning after three days of
heavy cloud, with dry, still weather when there
has been little or no significant change in the
ambient temperature! In those circumstances
the temperature of the (whole) structure will be
as close t o the ideal measured ambient air
temperature as it will ever be. As the
circumstances of this counsel of perfection are
seldom achieved, and even less likely t o be
possible within the programme constraints, the
advice is t o avoid the contrary situation, i.e. do
not attempt to set any critical components in
bright sunlight, on a windy day, or after a cold
night.
However, having made the above comments
about how bridges respond and when reliable
measurements can be made, it is pertinent t o
emphasise strongly that it will not be
necessary t o undertake an accurate exercise t o
establish the effective temperature of most of
the bridges that come within the range of
spans covered by this series of Guidance
Notes, i.e. up t o about 50 m.
Only for long structures, including multi-span
viaducts, and decks of more complicated
structures, such as lift bridges and swing
bridges, will it be necessary t o undertake
special measures to establish the effective
bridge temperature during the setting of
bearings, expansion joints and pivoting/ locking
devices.
Recommendations
In all cases, by far the best advice is t o
observe what happens at supports as often as
possible. A pattern of behaviour, related t o
conditions preceding and during observations,
is a much better guide than attempting t o
calculate behaviour in relation t o a limited
number of temperature measurements.
Correlating behaviour with predictions is much
more reliable in stable and overcast conditions
than in sunny or changing conditions, SO make
sure that there are sufficient observations
under stable conditions.
Ideally, all setting out activities that are
critically dependent on establishing effective
temperature should be carried out at about
7.02/2
© 2000 The Steel Construction Institute
Guidance Note
7.02
For most composite bridges (up to around
1 0 0 m length), very few temperature
measurements need t o be made in order to
establish a sufficiently accurate evaluation of
the position of the steelwork at 'datum'
temperature (and thus how the bearings should
be set). Measurements need only be made of
ambient shade air temperature in one or t w o
locations.
For longer bridges, more accurate evaluation
will probably be needed, but, fortunately, the
construction period will probably be longer and
this will offer more opportunity for observation.
It may be helpful to monitor steel temperatures
at the same time as ambient temperatures. In
such cases temperatures at top and bottom
flanges should be measured.
References
1. TRRL reports:
LR 561, The calculation of the distribution of
temperature in bridges, Emerson, M., 1973.
LR 696, Bridge temperatures estimated from
the shade temperature, Emerson, M., 1976.
LR 702, Bridge temperatures calculated by a
computer program, Emerson, M., 1976.
LR 744, Extreme values of bridge
temperatures for design purposes, Emerson,
M., 1976.
(All published by Transport and Road Research
Laboratory, Crowthorne.)
SCI-P-185 Guidance notes on best practice in steel bridge construction
7.0213
Steel Bridge Group
Guidance Note
Method statements
No. 7.08
Scope
The purpose of this Guidance Note is t o review
the use of method statements in the construction of steel bridgeworks. In particular, it gives
guidance on best practice for generation,
review and control of the definitive form of the
method statement used on site b y the bridge
contractor t o carry out the work. The quality
of that document is critical t o building the
bridge correctly in a safe planned manner.
Terminology
The term 'method statement' is used widely
throughout the course of a project, from
concept t o completion, t o refer t o a range of
quite different documents. For clarity in this
Note, the following terms are defined:
Bridge Contractor: the organisation , probably
a specialist sub-contractor, that is directly
responsible for erecting the bridgeworks.
Method statement: any document used in
some manner t o describe the erection method
during the course of a project, from concept to
completion.
Erection Method Statement: the Bridge Contractor's document that he uses for implementing the erection method.
The term 'Safety Method Statement' is used
in some HSE publications covering construction
generally t o describe a document used by a
contractor t o set out his safe system of work
for a construction activity.
As described
below, the Erection Method Statement covers
more than this.
Health and safety
The regulation of health and safety was rationalised in the Health and Safety at Work Act,
1974. Recognising that safety on construction
sites was heavily influenced by decisions in the
conceptual, detail design and procurement
phases of a project, the HSE published its
Guidance Note GS 28, Safe Erection of Strucrures (Ref 1 ) in 1984. For many years this set
out good practice for all parties t o a steelwork
project, and in particular it covered the need,
purpose and content of method statements in
general terms.
The regulation of health and safety has been
extended b y further legislation, particularly for
construction, the Construction (Design and
Management) Regulations, 1994. The CDM
Regulations placed the force of law o n owners
and designers, as well as contractors, t o have
due regard t o health and safety during construction, and for other phases of a project's
life from inception t o final demolition. (See
further comment in GN 9.01 .) The expectation
of good practice became a legal requirement.
GS 28 was withdrawn in 1997 for revision; a
replacement is due in 2001.
The following points are basic t o health and
safety considerations for the methods and
method statements for the erection of steel
bridges:
the designer of the bridge (as CDM defines)
has t o anticipate erection throughout, t o
ensure that erection is practicable and t o
minimise hazards and reduce risk as far as
practicable
the designer has t o communicate unusual
features and hazards, as well as his technical requirements, t o the Bridgework Contractor (through the supply chain)
the Health and Safety Plan for any bridge
project will require the Bridgework Contractor t o work t o documented safe systems of
work contained in a method statement
all designers, for permanent works, for
temporary works and for construction
engineering, are required t o cooperate with
regard t o health and safety.
Erection method
A n e w steel bridge is the product of the combined efforts of an owner and a set of designers and contractors. From concept t o completion, there is a simple sequence o f activities by
the participants in which erection is the culmination, if not the conclusion. Consequently:
the erection method is inextricably linked t o
the permanent works design
the method has t o be anticipated in all the
preceding activities
the choice of method determines much of
what goes before erection.
Clear communication about method is as
important as the drawings and the specification
- the better the communication, the better the
objectives of safety, economy and quality will
be met.
SCI-P-185 Guidance notes on best practice in steel bridge construction
7.0811
Guidance Note
No. 7.08
Method statements are used t o communicate
the method up and down the contractual chain,
for a variety of purposes throughout the procurement and construction phases.
This Guidance Note is primarily concerned with
the culmination of this process, the method
statement prepared by the Bridge Contractor
t o reflect all the requirements and constraints
of the contract, his o w n assessment of hazard
and risk, and his obligations under the Health
and Safety at Work Act.
The Bridge Contractor's Erection Method
Statement
Historically, bridge contractors' method statements have been technical documents with
explicit control of safety of the works, but only
implicit control of the health and safety o f
people.
In steel bridge building today, the Bridge Contractor's method statement has three essential
functions to fulfil in setting out explicitly the
plan for carrying out the work. The Erection
Method Statement has t o communicate:
1) clear instructions for site management and
responsibilities
2) engineering instructions t o site management
for the work necessary to achieve the
technical performance
3 ) the safe systems of work to undertake the
essentially hazardous tasks inherent in steel
erection.
Production of the Erection Method Statement
The Bridge Contractor is engaged in dialogue
about his method with the other parties from
the start of his contract: he also has t o carry
out his o w n design and planning for construction. Only when the method is agreed and his
design is substantially finished can the Erection
Method Statement be written ready for use on
site - and following on his o w n risk assessments.
The extent of the Bridge Contractor's design
will depend on the scale and complexity of the
bridge and will cover:
selection of equipment, plant and access
systems
*
resolution of the requirements of the contractors, utilities, and interested parties.
For this, the Bridge Contractor needs time, and
the Principal Contractor must ensure that this
is allowed for sufficiently in the Bridge Contractor's programme. The project programme
also has t o allow sufficient time for review of
the Erection Method Statement.
Reviewing an Erection Method Statement
In most projects that include steel bridgework,
the Erection Method Statement will be reviewed internally by the steel bridge contractor's originator and other managers (for engineering, safety or project considerations) and
by the main contractor (Principal Contractor)
the engineers responsible for the permanent
works (Designer) and for supervision of the
works (e.g. the Employer's Agent). Each
reviewer will bring his o w n specialist expertise
and particular project knowledge to bear on the
document: each has a contribution t o make by
way of constructive criticism in fulfilment of
his obligations under the health and safety
legislation. Note that these obligations cannot
be overridden by the terms of the contracts.
It is important that each party ensures that the
review is carried out by a competent person in
a co-operative and expeditious manner; the
purpose of the exercise is to enable the Bridge
Contractor t o implement his plan in the knowledge that it is sound.
What to look for in the Erection Method Statement
Faced with an Erection Method Statement for
review, ask the following questions of it.
Are the purpose and scope of the Erection
Method Statement clearly expressed?
what is covered?
what is excluded?
is it a controlled document from an effective quality management system?
choice of method
analysis of the structure for each stage;
design of temporary works
7.0812
© 2000 The Steel Construction Institute
Guidance Note
No. 7.08
Are the necessary and sufficient supporting
documents referenced?
Are the preparations for each stage of operation properly described?
are there meaningful sketches and drawings
of erection sequence and temporary works?
what equipment and plant are required?
what preparations are required by others?
what contract drawings and specifications
are required for erection?
are adequate contingency arrangements
provided for?
are crane duties documented?
policies
Are the instructions for each stage of operation
clear, explicit and unambiguous?
Is health and safety policy adeguately described?
is the contractor's safety policy invoked?
Is the Erection Method Statement complete?
are all safe systems of work covered or
identified for the site manager t o prepare?
does the Erection Method Statement antici-
what
project-specific regulations or
apply?
are special hazards identified (e.g. power
lines and hazardous products), and are
procedures t o deal with them in place?
who is responsible for safety on the site for
these works?
what evidence is there of a documented risk
assessment?
have the residual risks identified in the
health and safety plan been allowed for?
Is management o f the works clearly identified
and assigned?
who is in charge of the works?
who is in charge of each crane lift?
are responsibilities for interfaces and supporting or dependent activities defined?
(e.g. with the Main Contractor or Engineer's
Representative)
are there formal arrangements for coordination with all on site?
are handover or permit-to-work procedures
defined?
what engineering back-up is provided to
deal with unforeseen problems?
Are the site, the structure and the logic of the
scheme adequately described for a competent
site manager to understand the method, its
constraints and limitations?
Is the construction logic clear and sufficient?
are options allowed for, or is unnecessary
logic imposed?
are hold points and acceptance criteria
properly identified?
pate all known or possible hazards?
does it take account of any relevant matters
in the health and safety plan?
On a subjective level, there are sometimes
issues of style, undue brevity, superfluous
material and presentation. The originator
should be required t o address these only if they
are significant to the ultimate use of the document.
Having completed a review there are t w o
acceptance criteria that should be tested:
( 1 ) Is the Erection Method Statement, with its
reference documents, complete and sufficient for a competent site manager with no
previous information t o implement it as a
safe system of work? (It is not unknown
for personnel t o be introduced t o a project,
especially on small bridges, at a late stage.)
(2) if challenged, can the originator and the
reviewers demonstrate from the Erection
Method Statement how it satisfies all the
technical, safety and management requirements? (One could be faced with a lawyer!) A documented record of review/
comment is most effective in this regard.
Change control
The Erection Method Statement is finalised and
submitted for review near the end of the
contractor's design and planning work, so that
it will reflect fully the conditions under which
the work is done. It is inevitable, however,
from the nature of civil engineering construction that plans change - preceding work may
SCI-P-185 Guidance notes on best practice in steel bridge construction
7.08/3
Guidance Note
No. 7.08
be delayed, access may be lost after bad
weather, major plant may become unavailable
- in which case the method statement will
require revision, unless such change is anticipated by options in the text.
As for any other controlled document, change
to the Erection Method Statement would be
carried out by the originator and would undergo the same review process as before. This
may need t o be dealt with urgently: a change
can be required at the last minute, yet be a
very practical problem that needs understanding and co-operation t o expedite the solution
whilst maintaining the integrity of the construction process.
NOTE
The Erection Method Statement is a vital
document in bridge building; it is the Bridge
Contractor's document, but it requires the
whole project team's contribution to ensure its
validity; a large part of the value of preparing
and reviewing a Method Statement is acquired
during the process itself.
References
1. Health and Safety Executive, Guidance Note
GS 28, HMSO, 1 9 8 4 (withdrawn Dec
19971.
2. Health and Safety Commission, Managing
construction for health and safety: Construction (Design and Management) Regulations 1 9 9 4 Approved Code of Practice L54,
HSE Books, 1995.
7.08/4
© 2000 The Steel Construction Institute
SECTION 8
PROTECTIVE TREATMENT
8.0 1
Preparing for effective corrosion protection
8.02
Protective treatment of bolts
Guidance Note
8.02
Steel Bridge Group
Protective treatment of bolts
Scope
This note covers the various metal coatings
that are applied t o bolts used in bridgework,
and the practical aspects to be considered.
Zinc coatings used in bolt manufacture
Electroplated zinc coatings
The coating of zinc is applied by the electrolysis of an aqueous solution of a zinc salt. The
minimum local thickness typically used in
bridge construction is 8 microns.
Why apply metal coatings to bolts?
With the exception of structures of weather
resistant steel (WRS), the long-term corrosion
protection t o bolt groups in bridge structures
is given by the full coating system. (For WRS
steel structures the bolts, nuts and washers
should be of WRS material and are not given
any protective treatment, unless the steelwork
is painted for some reason.)
Bolts of grades higher than 8.8 are not normally electroplated because of the danger of
hydrogen embrittlement during the plating
process. It is understood that, as a precautionary measure, the Highways Agency requires
that any electroplated bolts are heat treated t o
drive out any possible entrapped hydrogen.
Threaded components are very difficult t o blast
clean effectively, and even more difficult to
metal spray effectively. The normal and recommended approach is therefore t o procure bolts
that are protected by metal coatings during
manufacture in order t o avoid blasting and
metal spraying of bolted joints in the assembled steelwork.
The metal coating provides primary protection
during construction until the rest of the coating
system is applied. (For a major structure, this
may involve a long period of exposure for the
metal coating.) Thereafter, the metal coating
provides additional protection, depending on its
thickness, in the event that the paint system
suffers local breakdown in the long term; it
also offers continuing protection t o the concealed surfaces of the bolts.
Beware of cadmium coatings
Cadmium plating was frequently specified and
used up to the early 1990s. It is now prohibited for health and safety reasons. Cadmium
is highly toxic if vaporised; this could happen
if a cutting flame or welding arc came into
contact with a cadmium-coated surface.
On earlier structures, even when zinc electroplated coatings were specified, it was common
for fabricators to seek and be granted a concession t o use zinc-plated bolts with cadmium
plated nuts, it being well known that such a
combination gave lower thread friction and
reduced tightening problems.
Extreme caution is therefore necessary when
dealing with any bolted connections made
before 1995, as cadmium may be present even
if the original specification suggests otherwise.
Hydrogen embrittlement results from the
absorption of monatomic hydrogen during the
plating process. After installation, and under
load, hydrogen atoms can migrate to dislocations in the crystal structure of the metal
where they can gather t o form gaseous hydrogen which generates very high levels of internal stress. This can lead t o a catastrophic
failure of the fastener. Such breakages can
occur hours or even weeks after tightening.
'
Chromate passivation is essential for all zinc
electroplated components. It is a process by
which the surface of the zinc coating is converted to extend its life. There are four levels
of passivation. If no passivation is applied zinc
salts of a powdery appearance will form on the
surface very quickly. This is known as white
rusting.
The Specification for Highway Works (SHW),
Series 1900 (Ref 1), states that chromate
passivation shall not be applied t o electroplated
bolts which are t o be painted. The reason
behind this is that many commonly used paint
systems will not adhere to zinc-coated surfaces
without some form of etching t o that surface.
The higher levels of chromate passivation can
resist such an etching process.
Zinc-plated bolts for bridgework are supplied in
the clear passivated condition which is the
lowest level of passivation. This will not resist
a subsequent etching process and should
therefore be deemed t o be compliant with the
SHW.
The thickness of the coating is such that no
special measures are required with respect t o
thread clearance.
SCI-P-185 Guidance notes on best practice in steel bridge construction
8.02/1
Guidance Note
8.02
The SHW calls for all zinc plating of bolts up t o
3 6 mm diameter t o be carried out in accordance with BS 3 3 8 2 (Ref 3). The maximum
bolt size covered by that standard is 314 inch
and therefore the maxium standard metric bolt
size covered is M16. The majority of bolts
used in bridgework are of larger diameter.
Despite the wording of the SHW, the common
approach for bolts over 2 0 m m diameter is to
supply t o BS 1706:1960 (Ref 2). Note that
the 1990 revision t o BS 1 7 0 6 cannot be used
as it was converted from a specification t o a
method of specifying.
BS 7371-3, Coatings on metal fasteners:
Specification for electroplated zinc and cadmium coatings (Ref 5), was published in 1993.
It covers bolt diameters up t o 64 m m which
comfortably covers the range of sizes normally
used in bridge construction. Surprisingly there
is no link between BS 7371 and BS 3382.
Specifying zinc electroplated bolts t o BS 7371
t o grade Zn8A would avoid the use of obsolescent codes.
An I S 0 standard on electroplating of bolts is
also due for publication and it may be some
time before the confusing array of codes in this
area settles down.
Sherardized coatings
Sherardizing is a diffusion process in which
articles are heated in close contact with zinc
dust. The process is normally carried out in a
slowly rotating and closed container at a
temperature in the region of 385°C.
For bridgework, Class 1 coatings to BS 4921
(Ref 4) are normally specified. This gives a
coating of minimum thickness 30 microns.
Sherardizing tends to be used mostly to protect
higher tensile
steels
(greater than
1000 N/mm2) t o avoid the risk of hydrogen
embrittlement
which can occur with
galvanizing or electroplating. It is also the
normal treatment for load indicating washers.
Note that sherardizing is only suitable for
protecting higher tensile steels if the method
of cleaning the bolts prior t o sherardizing is
mechanical. If, for grades 10.9 and above the
method of cleaning is acid pickling, there is a
risk of hydrogen embrittlement.
8.02/2
The resulting coating has a matt grey
appearance. Orange staining may become
apparent on sherardized coatings early in their
exposure, but this is not detrimental t o their
performance.
The thickness of the coating requires the nut
t o be over tapped to create sufficient thread
clearance. (See Appendix B of BS 4921 for
details.)
Passivating is not necessary.
(3) Hot dip galvanized coatings
Bolts and nuts are dipped in molten zinc and
then centrifuged t o remove excess zinc. Such
products are commonly referred t o as spun
galvanized. The minimum local thickness is
43 microns. The governing standard is BS EN
I S 0 1461:1999 (Ref 7).
Hot dip galvanizing provides the highest level
of corrosion protection as it gives a
considerably thicker coating than either
sherardizing or electroplating
Bolts are galvanized after threading. Nuts are
over tapped up t o 0.4 mm oversize t o create
thread clearance. This is achieved either by
galvanizing standard tapped nuts and then over
tapping after galvanizing, or by galvanizing the
nuts as blanks and then tapping them over size
after galvanizing. Although both approaches
result in an uncoated female thread, this will be
protected by the coating on the male thread
when the fastener is assembled.
There are no problems galvanizing bolts up t o
and including grade 8.8, even if acid pickling
is used t o clean the bolts prior t o galvanizing.
However, t o galvanize grade 10.9 bolts
satisfactorily mechanical cleaning must be
used.
High temperature galvanizing is now available
from some manufacturers.
The normal
galvanizing bath has a temperature of
approximately 450°C. However it has been
found that if the temperature is raised t o
approximately 550°C, a more even coating of
zinc is achieved. By careful choice of suitable
material and processing, manufacturers can
ensure that the high temperature galvanizing
process does not have any significant
retempering effect on the bolt. This process
© 2000 The Steel Construction Institute
Guidance Note
8.02
is not specifically covered in BS EN I S 0 1461,
but is n o w being recognised in international
standards currently in draft.
It has been found that HSFG bolts coated by
the high temperature process exhibit
considerably less thread friction during
tightening than conventionally galvanized bolts.
Nonetheless, the use of a lubricant during
tightening of bolts is still recommended.
Currently, grades up t o and including 10.9 can
be obtained in a high temperature galvanized
finish, and in sizes up t o 24zmm diameter.
The major fastener manufacturers have made
considerable investments in developing
improved methods of galvanizing. However,
extreme caution should always be exercised if
galvanized bolts are procured through
stockists, especially if the galvanizing is being
arranged by the stockist; the process used
must be identified reliably.
The British Standard produced specifically for
hot dip galvanized fasteners is BS 7371-6
(Ref 6) . In the case of fasteners, this new
standard should take precedence over BS EN
IS0 1461, which is a general specification for
hot dip galvanizing.
Passivation is not necessary on a galvanized
finish.
Tightening zinc-coated HSFG bolts
The tightening of zinc coated HSFG nuts and
bolts requires special care.
Zinc coated
surfaces tend t o bind under high interface
pressure; this phenomenon is known as galling.
Lubrication is essential t o avoid a high
proportion of bolt breakages in the latter stages
of tightening.
The most effective and economic lubricant is
tallow which for the best results should be
sparingly applied t o the leading threads within
the nut and the face of the nut that contacts
the washer.
Over application of tallow, for example dipping
complete assemblies in molten tallow, gives no
advantage and can create additional problems
in cleaning prior t o painting.
Some specialist bolt suppliers whose products
require consistent torque I tension relationships
during tightening apply wax-based lubricants
t o plated nuts under factory conditions. These
are often water soluble and can be readily
washed off after installation.
Some
manufacturers add a dye t o the wax coating t o
distinguish such bolts from untreated items.
Oils and greases should not be used for the
lubrication of HSFG assemblies as there is a
high risk of contaminating faying surfaces and
significantly reducing slip factors.
There is evidence that high temperature
galvanized HSFG bolts tighten satisfactorily
without additional lubrication and certainly with
many fewer problems than conventionally
It is nonetheless
galvanized fasteners.
advisable t o lubricate the bolts t o avoid
breakages during tightening.
Painting zinc-coated fasteners
Most paint primers will not satisfactorily
adhere directly t o zinc-coated surfaces.
T w o approaches are commonly used t o provide
a key for paint systems. The first is t o fix the
bolts in the structure, as supplied, and then
utilise a paint system for the bolted joints that
includes, as a first coat, an etch primer. The
second is t o T-wash the bolts before fixing.
T-wash is a solution containing phosphoric acid
and copper carbonate. The phosphoric acid
etches the surface of the zinc and the copper
carbonate produces a blue/black surface
colouration showing that the surface reaction
is complete.
This discolouration is n o t
absolutely uniform and it is not necessary t o
t r y t o achieve absolute blue/blackness, as long
as it is clear that the solution has been applied
over the whole area. Further application will
only remove more zinc than is necessary.
T-wash should not be applied after installation
as contamination of adjacent surfaces is
inevitable and the acid content may damage
them.
The data sheets f r o m
m o s t paint
manufacturers state that T-wash should be
brushed on, but this is impractical for
fasteners; the normal approach is t o batch dip.
However, care should be taken t o remove the
SCI-P-185 Guidance notes on best practice in steel bridge construction
8.02/3
Guidance Note
8.02
items from the solution as soon as they
discolour. If the items are left in the solution
for too long, the zinc will be stripped. It is also
very important t o rinse clean and dry the bolts
thoroughly.
Primers, other than etching materials, are now
available that will give satisfactory adhesion
directly on zinc plated surfaces. These provide
an attractive alternative t o T-wash, the use of
which now creates many problems under
environmental and health and safety
legislation. However, it is strongly advised
that adhesion tests are carried out to verify the
performance of such primers prior t o their use.
Treatment of plated bolts after installation
Normally the specification of the protective
system for HSFG bolted joints requires surface
preparation of the joint material by blast
cleaning. This should not be taken t o include
the bolts, nuts and washers, as they are not
'joint material'. It is less satisfactory t o blast
off the fasteners' coating and then reapply
coatings than simply t o degrease the fasteners
and apply the remainder of the coating system.
However, if the the fasteners have only thin
plating and there has been lengthy exposure
that results in corrosion, then blasting of the
fasteners may be required.
References
1. Manual of Contract Documents for Highway
Works, Volume 1 : Specification for
Highway Works, Series 1900, Protection of
steelwork against corrosion.
2. BS
1706:1960,
Specification
for
electroplated coatings of cadmium and zinc
on iron and steel (Withdrawn)
3. BS 3382: Parts 1 and 2:1961 Specification
for electroplated coatings on threaded
components.
Cadmium on steel
components. Zinc on steel components.
4. BS
4921 :1988,
Specification
for
sherardized coatings on iron or steel.
5. BS 7371-3:1993, Coatings on metal
fasteners. Specification for electroplated
zinc and cadmium coatings.
6. BS7371-6: 1998, Coatings on metal
fasteners. Specification for hot dipped
galvanized coatings.
7. BS EN I S 0 1461:1999, Hot dip galvanized
coatings on fabricated iron and steel
articles. Specifications and test methods .
Recommendations
The following table gives an indication of the
costs of the various types of zinc coating
relative t o that of the untreated fastener:
Coating
cost
Zinc electroplated
20%
Spun galvanized
30%
Sherardized
35%
Note: The total cost of bolts is usually less than 1%
of the cost of the structural steelwork.
Of the three processes, the most effective in
terms of corrosion protection is spun
galvanized.
It is recommended that spun galvanized bolts
be specified wherever possible.
0.0214
0 2000 The Steel Construction Institute
SECTION 9
9.0 1
OTHER TOPICS
Construction (Design and Management) Regulations
Steel Bridge Group
Guidance Note
9.01
Construction (Design and Management) Regulations
Scope
This Note gives guidance on the effect of the
CDM Regulations, which came into force on
31 March 1995, on the design and construction of steel bridges. Other publications (Refs
1, 2, 3 , 4, 6, 7, 8 ) should be consulted for a
more overall treatment of the CDM Regulations.
The Regulations
The Seventh EC Directive on Safety recognised
that sites expose workers t o high risks, and
that many accidents are caused by inadequate
coordination and unsatisfactory architectural
and organisational options or poor planning.
The Directive therefore required member states
of the EU t o adopt minimum requirements for
the better level of protection of workers whilst
avoiding imposing administrative, financial and
legal constraints that would restrict creativity
and development.
the contractor are competent t o carry out the
work on the project and that they have adequate resources available t o carry out the
work.
The requirement t o ensure competence should
not be taken lightly. It is clear that only an
experienced steel bridge engineer who has
relevant experience on site, awareness of
safety legislation and knowledge of risk assessment methods will have the ability t o be in
charge of a steel bridge design. It is illegal for
a client t o appoint an incompetent or inexperienced designer.
Competence of designers
A n important pair of regulations concern the
competence of designers (or contractors and
planning supervisors). Nobody should arrange
for a designer t o prepare a design unless he is
reasonably satisfied that the designer:
has the competence t o prepare the design
There are a total of 24 regulations in the
resulting legislation, the CDM Regulations. In
the context of this set of Guidance Notes, the
most important regulations are:
will allocate the resources needed t o comply with the requirements placed on him by
the CDM Regulations.
8
Competence of planning supervisor,
designers and contractors.
9
Provision for health and safety (i.e.
enough resources t o be provided).
13
Requirements on designers.
membership of a professional body
14
Requirements on planning supervisors.
Requirements relating t o the health and
safety plan.
familiarity with construction processes and
impact of design on health and safety
awareness of relevant safety legislation and
risk assessment methods
15
16
19
The Approved Code of Practice (ACOP, ref 1)
suggests that the steps required for the owner/
client/ main contractor t o be satisfied might
include enquiries as to:
Requirements on and power of principal
contractor (usually the main contractor).
Requirements and prohibitions on contractors (e.g. the steelwork contractor).
health and safety practices of the designer
with respect t o his design
*
skills and training of staff
time allocated for work
The Regulations impose specific duties on the
client, designer, planning supervisor, principal
contractor and other contractors. The duties
are t o be achieved through the medium of a
health and safety plan. (See reference 9 for
the standards relating t o the health and safety
plan and the health and safety file that are
issued by the UK highway authorities, under
the Design Manual for Roads and Bridges.)
Responsibilities of the client
The owner/ client/ main contractor has a
responsibility t o ensure that the designer and
technical facilities of the designer
method of communicating design decisions
w a y in which information on remaining risks
will be communicated.
The requirements on the designer are:
t o ensure the client is aware of his o w n
duties under CDM
t o ensure that his design includes adequate
regard for, and where practicable avoids
and combats at source, foreseeable risks t o
SCI-P-185 Guidance notes on best practice in steel bridge construction
9.01 /1
Guidance Note
9.01
health and safety of persons at work (or
affected by such work) caused by the
design during construction (including
amendment and demolition), during cleaning
and during maintenance, at any stage
t o cooperate with the planning supervisor
and other designers (e.g. in providing safety
and other information to be included in the
health and safety plan and the health and
safety file).
Designers
Through the life of a project, there will be a
number of different designers. For example,
on a Design, Build, Finance, Operate (DBFO)
project there might be:
a) the Highways Agency's designers, who
specify the design documentation and
produce the project specification
b) the feasibility designer, who identifies the
need for a bridge in the tender documentation
c) the preliminary designer, who produces the
Technical Approval Form and outline design
for tender estimation
d) the detailed designer (the person normally
thought of as the "designer")
e) specification designers (e.g. for a corrosion
protection system)
work. The various hazards listed below should
be considered to the extent which is appropriate t o the design stage reached. It is also vital
that the various interfaces between designers
are fully defined to ensure that all hazards are
identified and suitably addressedIn this respect, the Health and Safety Plan is
particularly important for the interface between
the detail designer and the contractor/ fabricator designers. However, each designer needs
t o ensure that any risks he has identified and
are not designed out are recorded and passed
on t o the other designers.
Example: top flange stability
As an example, consider the stability of the
top flange of the mid span of a conventional
composite plate girder deck during concreting
of the deck slab. A t the minimum, the detail
designer should consider using a wider top
flange t o reduce the risk of instability, indicate
the assumed construction sequence and identify the risk of top flange instability during
concreting.
However, the detailed designer is often best
placed, through his knowledge of the structure
(and particularly for a normal composite deck),
to design the temporary bracing.
f) the fabricator's erection designer
Thus, the detailed designer might:
g) the contractor's falsework designer.
a) only identify risk of top flange instability, or
The various designers will need t o have different competences and the various bodies (HA,
concession company, consultant, contractor,
fabricator) arranging for the designers t o
prepare designs will need t o ensure the designers have the required competences and resources.
The exact division of work between the various designers will depend on:
b) identify the bracing locations and loads t o
be resisted, or
c) fully design and detail the bracing.
The majority of the SBG feel that method (c)
should be used in general, for normal steel or
composite structures using normal erection
methods.
For more complicated erection
methods, the design of bracing might be better
left to the fabricator's designer who would
have a greater competence in erection design.
a) contractual arrangements
b) complexity of erection (where the role of
fabricator's erection designer may become
much more important)
c) whether the contractor wishes t o submit an
alternative design.
All the above designers must consider the risks
t o health and safety produced by their design
9.01 /2
Risk assessment
The risk t o health and safety produced by a
design feature has t o be weighed against the
cost of excluding that feature by:
designing to avoid risks t o health and safety
tackling the causes of risks at source; or if
this is not possible
© 1998, 2000 The Steel Construction Institute
Guidance Note
9.01
reducing and controlling the effects.
- the designer is t o assess the danger of
instability during construction
'Cost' includes an evaluation of fitness for
purpose, aesthetics, buildability, environmental
impact and maintenance costs, as well as the
initial construction cost.
A risk assessment consists of:
-
(4) Construction of deck cantilevers
-
an identification of hazards
an assessment of the risks associated with
each hazard.
General points
The designer must ensure that the structure
can be built; obtain advice from an experienced fabricator/erector if necessary.
If an alternative is proposed by the contractor (e.g. significant changes in construction
sequence or complete changes in design),
then the responsibility shifts t o the contractor's designer who must carry out his own
risk assessment.
Hazards
(This Section lists a number of hazards that
need t o be considered. However, each structure needs t o be considered individually; the
following list should not be considered as
stability measures must be specified, if
considered necessary.
the designer is t o indicate in outline how
this can be safely accomplished.
(5) Buildabilit y
- access for fabrication (particularly welding
of stiffeners)
- access for erection (e.g. welds, bolt placement and tightening)
-
inaccessible bolts are a risk t o health and
safety.
(6) Lifting points
- the designer must indicate allowable locations for lifting points on major elements
- the designer is t o specify laminar defect
-
checks where through thickness stresses
are induced
on plate I girders, the web/flange weld may
need local strengthening.
(7) Trial erection
exhaustive.)
- all the above hazards may be relevant
The hazards associated with the method of
construction (from the viewpoint of the detailed designer) include:
- generally, the fewer the spans (over other
(8) Conflict with existing traffic
(1) Compression flanges of girders
- the designer is t o indicate in outline the
requirements for temporary bracing necessary t o ensure stability at all stages of
erection
-
traffic, etc.), the less the conflict, but
consider the reduction in risk against the
increased construction cost
when there is a pier in a highway central
reserve, the risks are greatly reduced by
using temporary concrete barriers
- the designer is t o specify the assumed
- place abutments at some distance back
construction sequence
the designer is to consider the stability of a
from the road, t o give greater working
space
-
single girder when landed.
(2) Torsional restraint at supports
- the designer must draw attention to the
- consider hazards posed by large plant
working close t o traffic (e.g. tall items of
plant, excavators that can swing around)
-
need t o provide restraint until permanent
bracing is installed.
(3) Aerodynamic excitation of incomplete
structure (only relevant for larger structures)
consider specifying screens (e.g. as required
by Railtrack adjacent t o railways).
(9) Deck construction over a traffic route
consider a full route diversion
-
-
design with "piece-large" t o accommodate
limited possession times
SCI-P-185 Guidance notes on best practice in steel bridge construction
9.01 /3
Guidance Note
9.01
-
-
lift beams in braced pairs for immediate
stability
avoid site splices over traffic routes (if not
possible, place near the edge of the carriageway or over lesser-used railway track)
- use ‘Omnia’ planks or precast slabs as
-
-
- provide the soffit of a deck slab with cast-in
permanent formwork
- specify an enclosure system to protect the
stainless steel sockets from which working
platforms can be hung
traffic route if work will be over a live route.
Include consideration of relevant risks of
penetration of the enclosure in order t o
specify a required standard of resistance t o
- consider specifying additional headroom t o
penetration
- provide easy access for inspection
- provide jacking stiffeners for bearing inspec-
- consider hazards associated with working
allow for maintenance platforms.
( 14) Parapets, movement joints, bearings
over water (rivers, canals).
(10) Other hazards
- working at height
- confined spaces and difficult access in
-
consider provision of maintenance gantries
or walkways
use durable paints (particularly over traffic
routes) which can be easily prepared and
overcoated
tion and replacement
- specify durable products
- risks associated with aluminium parapets
are lower as no repainting will be needed.
boxes
( I5) Motorway widening
noise (e.g. use of mechanical equipment in
box girders)
- large risks associated with reconstruction
and discontinuous hard shoulders
- hot work (welding and thermal cutting)
- temporary loading of incomplete structure
-
- temporary works.
- consider designing bridge with long enough
consider whether reasonably foreseeable
that a motorway could be widened
spans t o allow widening in future.
The hazards associated with materials and
equipment used during construction include:
( 1 6) Box girders
-
( 1 1 ) Corrosion Protection
- see SA 8/94 (Ref 5) for details of hazards
associated with construction materials
- consider weather resistant steel (thus
avoiding risks associated with the application of paint)
- consider access during fabrication, erection
and maintenance
- provide drainage routes (unless sealed)
- provide access manholes, manholes in
- maximise shop painting.
( 12) Equipment used during construction
- minimum safe distance t o power lines
- location of underground services
- access of large cranes to site, including safe
bearing pressures at bearing pads.
The hazards associated with an operational
structure include:
( 13) Corrosion protection maintenance
-
consider weather resistant steel or enclosure systems
9.01/ 4
consider whether the box is t o be sealed
(with desiccant) or ventilated
-
diaphragms, etc. at safe locations and t o
required sizes for access
consider problems of confined space, e.g.
alternative methods of escape, ventilation.
ClRlA Report 145 (ref 6) gives a number of
case studies
Consequences of CDM on bridge construction
The likely effects of the CDM Regulations
are/should be:
more consideration of the competence of
designers rather than simply of the design
fee
© 1998, 2000 The Steel Construction Institute
Guidance Note
9.01
more systematic consideration of risks
associated with construction
an improved safety record in construction
and maintenance
increased use of weather resistant steel
increased provision of maintenance gantries
and walkways
increased use of precast concrete for decks
increased appreciation of latent risks by
personnel during maintenance, amendment
and eventual demolition.
References
1. HSC: Managing construction for health and
safety: Construction (Design and Management) Regulations 1994 Approved Code of
Practice L54, HSE Books 1995.
2. HSC: A guide to managing health and
safety construction, HSE Books 1995.
3. HSC: Designing for health and safety in
construction: A guide for designers on the
Construction (Design and Management
Regulations) 1994, HSE Books 1995.
4. Raymond Joyce, The CDM Regulations
Explained, Thomas Telford 1995.
5. The Highway Agency et al:
Design Manual for Roads and Bridges:
SA 8/94: Use of substances hazardous to
health in Highway construction, HMSO
1994.
6. ClRlA Report 145: CDM Regulations - case
study guidance for designers. An interim
report, ClRlA 1995.
7. ClRlA Report 166: CDM regulations, work
sector guidance for designers, ClRlA 1997.
8 . The Construction (Design and Management)
Regulations 1994: Advice for designers in
steel, The Steel Construction Institute,
1997.
9. The Highway Agency et al:
Design Manual for Roads and Bridges:
SD 10/95: The Construction (Design and
Management) Regulations 1 9 9 4 - Requirements for Health and Safety Plan
SD 11/95: The Construction (Design and
Management) Regulations 1994 - Requirements for Health and Safety File, 1995
(both published by:) HMSO 1995
SCI-P-185 Guidance notes on best practice in steel bridge construction
9.01 /5