Integral Bridges
Integral Bridges
Integral Bridges
Discuss me ...
Design Guidance
Published by:
The Steel Construction Institute
Silwood Park
Ascot
Berkshire SL5 7QN
Telephone:
Fax:
01344 623345
01344 622944
Discuss me ...
should
be
sent
to
Although care has been taken to ensure, to the best of our knowledge, that all data and information
contained herein are accurate to the extent that they relate to 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 to their use.
Publication Number: P163
ISBN 1 85942 053 2
British Library Cataloguing-in-Publication Data.
A catalogue record for this book is available from the British Library.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
ii
15 September 2003
FOREWORD
Integral bridge construction eliminates the provision of movement joints between
superstructure and substructure and thus avoids details that have given rise to many
durability problems in the past.
A number of studies have been carried out by The Steel Construction Institute on the
behaviour of integral bridge structures and this has led to the conclusion that the use of steel
elements in the bridge substructure (sheet piling, High Modulus Piles and steel bearing piles)
offers alternative construction sequences and methods which may well be cheaper and more
fit-for-purpose than the traditional reinforced concrete form of construction.
The purpose of this publication is to provide advice and guidance in the design of integral
bridges that use steel in a composite deck, in the substructure, and in both. It is also
intended to promote innovative thought by designers on alternative means of providing
bridge supports in integral bridges to those used traditionally in non-integral bridges. In
presenting new forms of substructure, the guide draws on technology that has been
developed over the past three decades in the Offshore oil and gas construction industry.
Use of steel in the substructure to bridges saves in dead load, provides material ductility and
permits speedier construction, all of which are significant advantages on many bridge
schemes. The use of prefabricated steel deck beams and steel piling saves site occupancy
time and minimises the traffic interruption for replacement bridge projects. It is hoped that
this Guide will encourage designers and constructors to consider a steel substructure option
more frequently during the conceptual and preliminary design phases of projects and thereby
to take advantage of the available potential to build more efficiently.
During the preparation of the publication, comment was received from the following people,
and their advice is gratefully acknowledged:
Mr
Mr
Mr
Mr
S G Griffiths
B Simpson
J L Vincett
R E Craig
Funding for the initial studies and for part of the cost of preparing the text of this publication
was provided by British Steel, Sections, Plates & Commercial Steels and by British Steel
Tubes & Pipes. The assistance of Mr W Ramsay, Mr J Wilson and Mr E F Hole of British
Steel is also gratefully acknowledged.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
iii
15 September 2003
Discuss me ...
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
iv
15 September 2003
CONTENTS
Page No.
SUMMARY
viii
INTRODUCTION
2
2
2
3
4
5
6
WHY
3.1
3.2
3.3
3.4
3.5
3.6
HISTORICAL BACKGROUND
4.1
Integral bridges in Europe and the USA
4.2
Examples of integral bridges in the United Kingdom
4.3
Offshore experience with tubular hollow sections
15
15
17
19
21
21
25
26
28
29
31
33
35
35
37
38
39
40
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
9
9
10
10
11
12
12
15 September 2003
6.6
6.7
6.8
41
41
42
DESIGN BASIS
7.1
General principles
7.2
Design standards
7.3
Limit state design
7.4
Loading effects on foundations
7.5
Observational Method for foundation design
7.6
Design report
7.7
Design for fatigue
46
46
46
50
51
51
52
53
DESIGN METHODOLOGY
8.1
Design sequence
8.2
Preliminary stages
8.3
Design of embedded retaining wall abutments
8.4
Design of column-pile abutments and piers
8.5
Design of bankseat integral bridges
8.6
Deck design
54
54
56
57
63
65
66
67
67
67
68
68
10
70
70
71
11
76
76
76
77
79
79
81
12
84
84
85
86
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
vi
15 September 2003
12.4
12.5
Requirements of BA 42/96
Comparison of Kp values in BA 42/96 with BS 8002
88
90
13
14
101
101
101
103
107
109
15
110
110
111
16
REFERENCES
112
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
vii
93
93
94
95
96
96
97
98
99
99
15 September 2003
SUMMARY
Integral bridge construction is now being actively pursued as a means to avoid
durability problems associated with the movement joints used in traditional
beam-type bridges.
This publication explains what is meant by an integral bridge and illustrates the
various structural configurations that may be used. A key aspect of the
performance of an integral bridge is that the bridge supports, and the soil that they
retain, are displaced by the cyclic thermal strains experienced by the bridge deck.
Bridge designers will need to learn how to deal with the response of the soil and
support structures to such displacements and to develop expertise in this new
concept.
Guidance is provided on the design basis for integral bridges and the design
methodology that will need to be followed, both for bridges with retaining wall
abutments and for bridges on bankseats or individual pile supports.
The use of steel piling in the bridge supports offers a compliant structural element
that is well suited to integral bridge construction. The behaviour of the steel
supports under the loads from the deck and pressure from the soil is explained.
The requirements of the Highways Agency are discussed and compared with other
standards and design rules relating to soil behaviour. The interaction between the
stiffnesses of the deck, the supports and the soil is explored and the requirements
for the connection between the two are examined.
Reference is made to the companion publications Steel integral bridges: Design of
a single-span bridge - Worked example and Steel integral bridges: Design of a
multi-span bridge - Worked example, which illustrate many of the aspects covered
in this publication.
Pont en acier de type intgral: guide de dimensionnement
Rsum
La construction de ponts de type intgral est actuellement en plein essor car elle
permet dviter les problmes de durabilit lis aux appuis mobiles des ponts
poutres traditionnels.
La publication explique le concept de pont intgral et illustre les diffrentes
configurations structurales qui peuvent tre utilises. Un point trs important qui
conditionne le bon comportement de ce type douvrage est celui des dplacements
provoqus, dans les appuis et le sol quils retiennent, par les mouvements du
tablier du pont dus aux variations thermiques. Il est indispensable que lingnieur
projeteur soit bien au courant de ce problme et puisse le prendre en compte de
manre correcte.
Le guide couvre les points principaux du dimensionnement des ponts de type
intgral et expose la mthodologie suivre tant pour le dimensionnement du pont,
pour les murs supports situs aux extrmits du pont et pour les piles de ponts.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
viii
15 September 2003
Zusammenfassung
Der Bau von Rahmenbrcken wird aktiv verfolgt als ein Mittel, Probleme der
Dauerhaftigkeit zu vermeiden, die sich bei gewhnlichen Balkenbrcken infolge von
beweglichen Auflagern ergeben.
Diese Verffentlichung erklrt den Begriff Rahmenbrcke und zeigt die
verschiedenen statischen Systeme. Ein entscheidender Gesichtspunkt des Verhaltens
einer Rahmenbrkke ist die Verformung im Bereich der Auflager und des gesttzten
Bodens infolge zyklischer, thermischer Dehnungen des Brckenbalkens.
Brckenplaner mssen lernen, mit der Antwort des Bodens und der Auflager auf die
Verformungen umzugehen und Erfahrung mit diesem neuen Konzept zu sammeln.
Grundlagen zur Berechnung von Rahmenbrcken und die anzuwendende
Berechnungsmethodik werden vermittelt, sowohl fr Brcken mit Widerlagerwnden
als auch fr Brcken mit Auflagerbnken oder Auflagem aus Pfhlen.
Stahlpfhle fr die Brckenauflager sind ein gnstiges bauliches Element, das gut
zum Bau von Rahmenbrcken pat. Ihr Verhalten unter der Belastung aus dem
Brckenbalken und dem Erddruck wird erklrt.
Die Anforderungen der Straenbaubehrde werden besprochen und mit anderen
Vorschriften und Berechnungsregeln hinsichtlich des Bodenverhaltens verglichen.
Die Interaktion zwischen der Steifigkeit des Brckenbalkens, der Auflager und des
Bodens sowie die Anforderungen fr die Verbindung zwischen den beiden, werden
untersucht.
Auf die begleitenden Publikationen Rahmenbrcken aus Stahl: Berechnung einer
einfeldrigen Brcke - Berechnungsbeispiel und Rahmenbrcken aus Stahl:
Berechnung einer mehrfeldrigen Brcke - Berechnungsbeispiel wird Bezug
genommen; sie illustrieren viele der Aspekte, die in dieser Verffentlichung
behandelt werden.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
ix
15 September 2003
Viene presentata una guida per la progettazione di base di ponti integrali, con
riferimento alla metodologia di calcolo da utilizzare, sia per ponti con spalle a
parete sia quelli che poggiano su argini o su singole pile.
Luso di pile in acciaio per lappoggio della travata rappresenta una soluzione
estremamente conveniente, bene integrabile con il sistema costruttivo in esame. E
analizzato il comportamento degli appoggi da ponte in presenza dei carichi
trasmessi dallimpalcato e delle azioni esercitate dal terreno.
I requisiti di queste strutture imposti dagli enti preposti alla viabilit sono discussi
e paragonati con altri criteri generali e con regole di dimensionamento legate al
comportamento del terreno. Linterazione tra la rigidezza di impalcato, appoggi
e terreno analizzata e sono esaminati i requisiti dei collegamenti.
Viene fatto riferimento alle pubblicazione sulla stessa tematica Ponti integrali in
acciaio: progettazione di un sistema a campata singola - esempio applicativo e
Ponti integrali in acciaio: progettazione di un sistema a pi campate - esempio
applicativo, le quali trattano molti degli aspetti affrontati in questa guida
progettuale.
Puentes de acero integrales: Gua de Proyecto
Resumen
Actualmente la construccin de puentes integrales se ve favorecida con un intento
de evitar los problemas de durabilidad asociados al movimiento de las juntas
tradicionalmente utilizadas en los puentes de vigas.
Esta publicacin explica lo que se entiende por puente integral e indica las
tipologas estructurales utilizadas. Un aspecto clave en el funcionamiento de un
puente integral es el desplazamiento impuesto por las deformaciones trmicas
cclicas del tablero a los apoyos y al suelo retenido por aquellos. Los proyectistas
de puentes debern familiarizarse y hacerse expertos en el tratamiento de la
respuesta del suelo y las estructuras de soporte ante aquellos desplazamientos.
En la obra se dan indicaciones sobre las bases de proyecto de puentes integrales
as como sobre la metodologa que debe seguirse tanto para puentes con estribos
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
de contencin como para puentes con soportes tipo durmiente o con pilares
individuales.
El uso de pilas de acero en los apoyos permite disponer de un elemento flexible
muy adecuado para la construccin integral. Su comportamiento bajo las cargas
del tablero y de la presin del suelo se explica cuidadosamente.
Tambin se analizan los requisitos establecidos por la Highway Agency,
comparados con otras Normas y Reglas de buena prctica relativas al
comportamiento del suelo.
Se estudia la interaccin entre las rigideces del tablero, soportes y suelo, as como
los requisitos de unin entre aqullos.
A lo largo del trabajo se hace referencia a la publicacines gemelas tituladas
Puentes integrales de acero: ejemplo desarrollado para un puente de un vano e
Puentes integrales de acero: ejemplo desarrollado para un puente de varios
vanos que en una serie de hojas de clculo ponen de manifiesto muchos de los
temas contenidos en esta obra.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
xi
15 September 2003
Discuss me ...
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
xii
15 September 2003
1 INTRODUCTION
A modern beam-type bridge comprises two essential structural components - a deck
to span the gap and the supports. The deck has roadway surfacing, and that
surface must match against the surfaces of the approach roads at either end. Since
these bridge structures flex, expand and contract, it has been customary to use
separation joints between the ends of the deck and the approach structures, and to
provide simple bearings on the supporting structures. The object has been not to
constrain the thermal expansion or contraction of deck beams nor to restrain their
rotation at supports.
Multiple span bridges can then be formed from a series of simply-supported beams,
with similar separation joints between the ends of the separate decks - this was
particularly popular in construction in reinforced concrete. However, structural
continuity over intermediate supports has always been easy to achieve with
composite decks, and this has little effect on the support structure, other than a
redistribution of the vertical reactions, but it does afford some economy in the deck
construction.
The consequence of providing simple support details and separation joints is that
the abutting interface between bridge road surface and approach surface sees a
range of movement as the bridge temperature changes. For very small bridges this
can be accommodated by a narrow gap that opens and closes, but for larger bridges
a fabricated movement joint must be provided, so that the gaps are never large
enough to cause a hazard to the road users.
The structural form of a beam-type bridge with movement joints may be contrasted
with that of a traditional masonry arch. The arch will change its shape slightly
under load as it springs load in compression to the abutments, but the roadway
is effectively continuous, laid on approach road base foundations and then on fill
over the arch barrel. Such structural deformations as occur are accommodated
within the fill, road base and surfacing materials. With a masonry arch bridge,
there is no gap, no discrete interface, no relative movement between the bridge
roadway and the approach roadway, because the arch, its abutments and the soil
behind all act together, or integrally.
With beam-type bridges there have been many problems in practice with leaking
joints, both over intermediate supports and at end supports, leading to poor
durability and consequent high maintenance costs. As a result, the Highways
Agency (HA), would like to see greater use of integral construction i.e. without
movement joints, particularly for bridges shorter than 60 m.
This publication is based on findings from studies carried out by the SCI for British
Steel(1); it provides an introduction to the concepts relating to integral bridges and
illustrates ways in which the ordinary composite beam-and-slab deck bridge can be
adapted to become an integral bridge. Also, the opportunity to use steel in place
of reinforced concrete for the supports is explored. Steel piles offer a degree of
flexibility at supports that is particularly suited to the movements that occur in an
integrated structure; guidance is included to facilitate the consideration of steel
piled substructures.
Reference is made throughout to two companion
(2)(3)
that
illustrate
by worked examples many of the design aspects
publications
covered in this publication.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Generally, an integral bridge is one where the bridge deck is made without any
movement joints at the abutments or between spans. The use of the terms integral
bridge, integral abutment and integral construction have not been consistent to
date and the extent of the integrity between the deck and supporting structure
varies. To avoid confusion a more rigorous system of definition is required.
Bridges without movement joints can be conveniently divided into two basic classes,
termed integral, and jointless deck. The difference between the two, and the
principal features of various forms of each, are explained below.
The use of the term integral abutment is avoided in this publication, except when
referring to its use in the USA.
The term endscreen, or endscreen wall, is used in relation to both integral and
jointless deck bridges to describe the stub wall at the end of the deck that retains the
adjacent road construction.
2.1.3 Supports
Integral bridges can have wall abutments, column piers or bankseat supports or
combinations. The foundations can be either spread footings or piles. Conventional
abutments comprise retaining walls where either concrete types or sheet piles are
used. Other types of abutment include pier abutments that are essentially
fall-through column abutments with endwalls and side slope configurations.
2.2
The frame abutment integral bridge is a fully integral bridge with the abutment walls
working integrally with the soils that surround them (and thus derive some of their
resistance from them to lateral loads in bending). In addition, the supporting
elements carry the axial loads which are the end shear forces from the deck beams.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Sway restraint is provided by the soil between the 'foundation' and the top of the
frame abutment but the degree of restraint is dependent on the soil characteristics
and the geometrical configuration of the supporting element.
To illustrate the effects that must be considered in the design of the fully integral
bridge, a diagrammatic illustration of the deflections due to a temperature increase
and due to live loading on the bridge are given in Figures 2.1 and 2.2. No
particular form of foundation is shown in order that the diagrams can be taken to
represent either a wall on a strip or spread footing, a wall on a pile cap foundation,
or the upper part of piles driven to a greater depth.
Figure 2.1
Figure 2.2
2.3
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Figure 2.3
2.4
Figure 2.4
Figure 2.5
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
A key aspect to note is that the foundation face which bears on the ground can slide
on the soil under certain combinations of loading (that include significant thermal
strain) and may also rock as the deck deflects under live loading. Depending on the
soil type, the soil may be affected by these cyclic movements, and the possibility of
degradation of the bearing strength needs to be taken account of in deciding a
permissible bearing pressure. Long term settlement could result beneath such bridge
ends. Therefore bankseat integral bridges should only be used where the soils have
high strength, and the total length of the bridge is small (short bridges have lesser
end movements and rocking under load).
2.5
Sliding bearing
Figure 2.6
Sliding bearing
Figure 2.7
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
2.6
Where the bridge deck length is small, thermal movements can be accommodated
at the bridge ends by an asphaltic plug joint in the road surface immediately behind
the end of the bridge. The joint can be expected to perform elastically without
cracking. However, when deck lengths and movements are larger, cracking is likely
to develop in the road surface in the joint area, allowing salt-laden runoff water to
percolate down the back of the structure.
Opinions vary as to the effectiveness of asphaltic plug joints but the prudent bridge
designer should pay attention to detail at the buried end of the deck and make
provision for surface water leakage regardless of the claims about so-called elastic
behaviour of joints. The difference in vertical stiffness between the road and the
end wall will concentrate any movement at the junction. Consequently, the surface
run-off on the road will doubtless find its way through cracks to the back of the
abutment wall. There is a concern for the maintenance authority that any
deterioration that results will be entirely hidden and un-inspectable.
For bridge lengths in excess of about 10 metres it is therefore perceived to be
difficult to produce a totally satisfactory and durable joint at the junction of the
endscreen wall and the road construction.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Road surface
Road base
Approach slab
Select granular
fill
Integral
bridge
deck
Embankment fill
Steel bearing
piles
Figure 2.8
The maintenance problem at the joint is transferred from the bridge structure to the
road construction with an approach slab, but it is possible that two structural
junctions are created at the changes in stiffness between deck/approach slab and
approach slab/road construction. Due to the likelihood of cracking at either
junction, it is considered that provision must be made to collect any surface water
that could percolate through such cracks at both, and to lead it away from the
adjacent bridge structure.
Any leakage at the movement joint can be captured by transverse drains at road
formation level and led away to the normal carriageway drainage system that is
remote from the structure, but careful attention to detail of the bridge end shape is
necessary to make such provision really effective.
Approach slabs have not been favoured by the Highways Agency for use in nonintegral bridges because of a poor track record. Their preference seems to be to
concentrate on the use of controlled backfills to bridge end-walls and attention to
careful compaction to prevent settlement. Whether this proves to be effective for
integral bridges remains to be seen.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Alternatively, if the approaches are not retained and have natural earthworks slopes
or reinforced earth slopes instead, column-pile integral bridges and jointless bridges
can be used with bankseats on footings or on piles as already described.
For motorway or trunk road replacement bridges, individual steel bearing piles or
sheet pile walls can be driven through the existing embankment fill or cutting face
to found on deeper soils before any new earthworks are started.
For new integral bridges on green-field sites, steel column-piles can be driven to
provide an upstanding length of column as an alternative to conventional concrete
construction; this can save site occupancy time.
In the preceding definitions and Figures, the bridge is taken to be a single span for
simplicity. Many bridges, however, are of more than one span. Clearly, to be
integral, the deck beams must be continuous over any intermediate supports, but
this arrangement is already quite normal for composite construction. However,
there remains the question of whether the behaviour of intermediate supports has any
effect on the overall integral behaviour.
If the intermediate supports are provided with sliding bearings offering no rotational
restraint, then clearly they will have no effect on the deck behaviour due to
temperature change. If the supports are pinned to the beams (but still with no
rotational restraint) - perhaps in order to offer restraint to the support against
collision loads - then they will be displaced by deck thermal strains unless it is the
central support to a symmetrical bridge configuration. This displacement is shown
in Figure 2.9.
Generally, thermal movements will be proportional to the distance from a null
point in the middle of the bridge, but if the spans are unsymmetrically disposed, or
if the stiffnesses at the ends are unequal, the null point will not be central.
Figure 2.9
Intermediate columns that need to be designed to resist vehicle collision loads can
be more effective when they are restrained at the top by the bridge deck, perhaps
through a crossbeam. Unless they are at the null point, the deflections due to
thermal movements will then have to be taken into account in the column design.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
Over the past thirty years, engineers have become more aware of the pitfalls
associated with the use of expansion joints and expansion bearings.
C
Joints are expensive to buy, install, maintain and repair; repair costs can be as
high as replacement costs.
Even so-called waterproof joints will leak over time, allowing runoff water,
often salt-laden, to penetrate through the joint and thus accelerate corrosion
damage to girder ends, bearings and supporting substructures.
Accumulated dirt, stones and rubbish may fill recesses, which can, for example,
lead to failure of elastomeric bearing glands.
Hardware for joints can be damaged and loosened by snow ploughs and heavy
traffic.
In time certain types of steel bearing may tilt and/or seize up because of loss
of lubrication or build-up of corrosion.
Seized expansion joints and malfunctioning expansion bearings can also lead to
damage of the main structural members.
In 1985 in the USA, a survey carried out by the Federal Highway Administration
found(6) that 75% of the bridges built using expansion joints and bearings
experienced movement contrary to their designers intent. The survey report pointed
out that vertical movements were noticeably greater than horizontal movements,
where the magnitudes of these vertical movements in many instances were due to the
inward movement of the abutments.
In the UK, a survey was carried out by Maunsell and Partners for the Department
of Transport in 1989. The report of that survey(7) identified a number of factors
which contributed to the inadequate durability of many bridge structures. The most
serious sources of damage were found to be salt water leaking through joints in the
deck or service ducts and poor or faulty drainage systems. Also, damage occurred
due to splashing or spraying of salt water from de-icing salts on to bridge abutments,
piers, parapet edge beams and deck soffits. Poor workmanship was found to be an
extremely frequent problem. Most critical was the failure to achieve the specified
concrete cover to steel reinforcement. This led to deterioration, particularly when
it occurred in association with joint leakage. Cracking was the other main problem,
particularly that due to early thermal effects.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15 September 2003
3.2
Based partly on the findings of the Maunsell report, the Highways Agency published
their Standard, BD 57/95 Design for Durability and the accompanying Advice Note
BA 57/95(8). BD 57 specifies that bridges with lengths not exceeding 60 m and
skews not exceeding 30E shall ... be designed as integral bridges ... without
movement joints for expansion or contraction.
That Standard is concerned mainly with the Agencys principles of design; details
of the Agencys advice on the design of integral bridges are given in their Advice
Note BA 42/96 Design of Integral Bridges(9) that was published in November 1996.
See Sections 7 and 12 for further discussion on BA 42/96.
3.3
Clearly, the first advantage is the elimination of the cost of, and additional work
associated with, the provision of movement joints at the ends of the bridge. This
advantage is confirmed by experience in the USA (see Section 4), where it was
found that the initial capital costs of integral bridges were cheaper than bridges with
expansion joints, even when the extra work associated with ensuring structural
continuity were taken into account (see Burke(10)).
The benefits of reduced maintenance costs and reduced risk of damage arising from
leaky joints is less quantifiable, but is probably the major benefit in most cases.
Apart from these two principal advantages, other benefits can be seen, depending
on particular circumstances and configurations. These include:
Substructure design
The restraint to retaining wall abutment structures provided by the deck (which can
act as both a prop and a rotation restraint) can lead to economies in the wall design.
Resistance to accidental and seismic loadings
The increased longitudinal restraint to the deck, and in frame abutment integral
bridges, the moment restraint, provide extra load paths against the effects of
accidental and seismic events. In particular, where seismic loading is a significant
consideration, considerable savings can be achieved by avoiding the need for
enlarged bearing seat widths and restraining devices.
Torsional restraint of deck at supports
Substantial endscreen or abutment walls ensure that all the deck beams, and the full
width of wall, rotate equally and thus tend to distribute loads more evenly between
the deck beams.
Faster construction
With piled abutments, only vertical piles are needed (no rakers), which both
simplifies and speeds construction. Where permanent bearings are omitted, a timeconsuming operation is eliminated.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
10
15 September 2003
3.4
The main practical factors influencing the choice of an integral configuration are:
C
The geometrical arrangement of the end supports (e.g. whether there are side
slopes, cut or fill earthworks).
Where the bridge is a replacement, or is built over an existing road, speed and ease
of construction will also have an important influence on the choice. This is
particularly so where lane rental charges apply and traffic disruption cost has to be
considered. Advances in steel bridge design and construction are providing bridge
engineers with deck structures that can be fabricated to a large extent off-site and
therefore require less time for erection. It is therefore logical to develop any
applications of steel for the support elements of the integral bridge that can also
minimise site occupation time.
When electing for an intermediate support to the bridge, the need to make use of the
deck to provide restraint to the top of the support to assist in resisting collision loads
is an important factor in detailing the structural framing and the connections. The
suitability of concrete, steel or composite construction at the connection will be an
important consideration at all supports. In addition, a more economic substructure
may be possible by considering compliant steel piled supports.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
11
15 September 2003
3.5
The savings in whole life cost by choosing an integral bridge concept can be:
C
Reduced allowance for the maintenance cost for the deck slab (because the
worst potential source for deck reinforcement corrosion is eliminated).
3.6
Steel piling can be used for elements in substructures and foundations for integral
bridges, such as:
C
Bridge abutments.
Intermediate piers.
Wing walls.
Retaining walls.
Steel pile sections have been used successfully in non-integral bridges in the UK and
elsewhere in the world and provide a prefabricated, high quality foundation of
known structural integrity that fulfils the requirement for minimum construction
time. Not only can piles be driven rapidly in the vast majority of soil types but they
are capable of being loaded immediately, which is a distinct advantage in fast-track
construction projects.
The advantages of steel piling are described in the Steel bearing piles guide(12) and
may be summarised as the follows:
C
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
12
15 September 2003
Sheet pile walls provide a curtain walling to contain the working site.
Frame
Pinned
Carriageway
Carriageway
Retaining wall
Retaining wall
Figure 3.1
Bankseat
Piled bankseat
Some types of jointless bridge with steel substructures are illustrated in Figure 3.2.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
13
15 September 2003
Cantilever/Propped
Bankseat
Sliding
support
Flexible seal
H piles
Carriageway
Retaining wall
Figure 3.2
Surface textures and paint colour of the facing walls, and the expression and
position of vertical and horizontal construction joints.
The appearance of a pile can be improved by providing features in the finished face
or by decorative facings or claddings.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
14
15 September 2003
4 HISTORICAL BACKGROUND
4.1
Bridges where joints within the deck of the bridge are omitted are not new concepts
in design. In some countries it has been usual practice to build bridges without
joints, and bridge spans longer than 100 m have been built. There are now
examples of modern integral construction in many countries including Australia and
Sweden, but the country with the greatest experience to date is the USA. Only a
few integral bridges have been built in the UK.
In Sweden, bridges are built without expansion joints and even without
transition/approach slabs. Various integral bridge types are in use without special
abutments in the embankments. Single-span slab-frame integral bridges have been
used for over 50 years and the longest continuous slab-frame bridge was one built
in 1968 which had five spans and a total length of 120 m. Sweden has the
advantage of rock foundations that permit the generation of stiff reactions to restrict
bridge movements.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
15
15 September 2003
Reinforcement
Prestressed
concrete beams
Selected
granular fill
Bearing
Drainage
aggregate
H piles
Figure 4.1
The introduction of rotational continuity can generate high bending stresses in the
abutment detail, and this is either accepted or Freysinnet hinge details were
introduced into the reinforced concrete wall. Examples of severe cracking and
splitting have been documented at Freysinnet hinges and the long term durability
of the integral abutment details used is not known.
The use of coated reinforcement bars appears to be normal procedure for such
abutments in USA now because they are concerned about the durability aspects of
spalled concrete cover. However, the design life for bridges in the USA is only 50
years(14). Consequently, care should be exercised before adopting any reinforcement
details or connection details from US practice.
Figure 4.2
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
16
15 September 2003
of how connections work and of the stress distributions involved than is presently
available in the USA.
Many lessons can be learnt from the US experience. Burke(4) has identified a list of
design recommendations to assist an engineer considering an integral bridge design.
These have been considered in Section 3 of this Guide.
Skew integral bridges
A nationwide survey in the USA has shown that skewed and curved integral bridges
where the deck is rigidly connected to the supporting wall structure are common,
and Greiman et al(17) summarized the findings of a survey of the Highway
Departments of all 50 States to obtain information on the design and performance
of skewed bridges with integral abutments. It was found that there was a lack of
theoretical and experimental research in this area(18)(19), with the result that most
states designed integral abutments on skewed bridges on the basis of empirical
experience without adequate theoretical analysis.
4.2
Portal bridges with monolithic abutments have been built in Britain - perhaps the
best known are the twin span concrete overbridges on the original section of the M1
motorway, designed by Sir Owen Williams & Partners in the early 1960s. This
type of construction is particularly massive and was not adopted on later motorways.
Recent examples of integral bridges include the Stockley Park Canal bridge near
Heathrow Airport; the A41 Stone Bridge, Aylesbury, Buckinghamshire; the Chad
Brook Bridge in Suffolk; the Bridgend-By-Pass overbridge; that all used steel piled
supports. Several others are currently being constructed.
The above named bridges all have reinforced concrete decks, but their features
nevertheless illustrate general principles that are applicable to both concrete and
composite deck bridges. The principle features are described below.
Stockley Park Canal Bridge
The Stockley Park canal bridge (Figure 4.3) has a clear span of 19 m and was
designed as an integral bridge such that the mid-span bending moment was a
minimum, in order to provide adequate canal traffic head clearance without raising
the road elevation. The deck is monolithic with the reinforced concrete abutment
walls, and live loads are carried by transferring moments into the abutments. The
abutments are founded on a single line of 600 m diameter steel tubular piles that are
embedded in the abutment retaining end wall. This has been designed by including
the rotational stiffness of the abutment fill in the overall stiffness of the bridge
structure. The design concept assumed that the abutment fill not only acted as a
load on the abutment but also provided an additional component of restraint in the
bridge structure. This example of an integral bridge demonstrates an elegant form
with construction economies. Further information is given in a paper by Low(20).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
17
15 September 2003
C
L
19m
CL Canal
Figure 4.3
Figure 4.4
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
18
15 September 2003
Road level
Extent of
bank facing
2
Box piles
forming portal
Figure 4.5
Bridgend-By-Pass
The Bridgend-By-Pass railway bridge is a twin 25 m span replacement frame
abutment integral bridge on steel sheet pile abutments and with an H-piled central
reservation pier built for Railtrack plc. The abutments are constructed in Larssen 6
sheet piles of approximate length 22 m. A reinforced concrete capping beam
provides the connection between the sheet piles and the reinforced concrete bridge
deck.
4.3
Structural steel circular hollow sections have been used extensively in the UK
offshore oil and gas industry over the last 25 years. During that time, a technology
has been developed which, through experimental research and testing, has produced
industry accepted practice and enabled codes, standards and guidance to be written
and comprehensive design procedures to be developed. The leading codes of
practice have been the American Petroleum Institute codes of practice (API) which
are being updated constantly to embody technical developments in the Oil and Gas
Industry. The code of practice relevant to steel structures is API RP2A(22) which
covers all aspects of design and construction, thereby enabling international design
consultants, fabricators and installation contractors to work to a common standard.
Extensive information is available on circular hollow sections because they are
chosen for the major structural components of offshore platforms including the main
support columns and legs, bracing and piling. Multi-million pound research
programmes have enabled comprehensive tests to be performed to study and simulate
the behaviour of tubular frames. This has included tests to analyse material
strength, welding, durability and fatigue aspects specific to the environmental
conditions that are present during their in-service life.
Over the past three decades, the offshore industry has invested heavily in R & D in
tubular steel piling technology. This has included:C
Improved soil investigation methods and tools, e.g. Dutch cone CPTs.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
19
15 September 2003
This has resulted in consensus amongst practising experts in the profession as to the
best geotechnical design methods to use for tubular steel piles. The methods have
a sound theoretical basis but the limitations of theory are recognised by including
empirical adjustment factors. These methods have been validated by the SCI and
are presented in the Steel bearing piles guide(12).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
20
15 September 2003
5.1
As explained in Section 2.2 this type of bridge is essentially a simple portal frame,
and this means that the end supports will be subject not only to bending due to
pressure of retained soil and axial load as a result of supporting the deck, but also
to additional bending as a result of increased soil pressure when the deck expands
and contracts thermally, and to moments and longitudinal forces transmitted to the
top of the wall from the deck.
Initial considerations by the SCI indicated that it would appear difficult to adapt the
traditional reinforced concrete wall-type abutment to accommodate significant
thermal displacements and the reverse curvature imposed on a stiff wall. For frame
abutment integral bridges the substructure elements need to be compliant in
accommodating displacements due to thermal strain, and steel construction is well
suited to act in this manner. Consequently, much of this publication concentrates
on the use of steel supports, primarily for frame abutment and pinned integral
bridges, but also for other types of support, e.g. steel column-piles.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
21
15 September 2003
Apart from the design of the wall section to carry the forces, the most important
aspect of the substructure to this type of bridge is the analysis and design of the
moment connection between the deck and the pile wall abutment. Further
information is given in the companion publications Steel integral bridges: Design of
a single-span bridge - Worked example(2) and Steel integral bridges: Design of a
multi-span bridge - Worked example(3).
Concrete capping beam connection
The most versatile method of connecting the deck to a pile wall is a reinforced
concrete pile capping beam. This is effectively an adaptation of a traditional r.c.
capping beam used with sheet pile walls in order to enable moments to be
transmitted. The in situ form of construction accommodates the practical tolerances
that have to be allowed for in the positioning and alignment of the pile wall
installation and for placement of the deck beams. Capping beam connections can
be used with both composite steel/concrete and in situ/precast concrete beam deck
construction.
A moment connection for a High Modulus Pile wall is shown in Figure 5.1.
Asphaltic
plug joint
Concrete deck
Deck
beam
Reinforced
concrete pile
capping beam/wall
Construction
joint
Note:
Similar arrangement
for reinforced
concrete deck beam
High Modulus
Pile
Sheet
pile
Figure 5.1
UB
Effective transfer of moment from deck beams is possible, provided that the concrete
capping beam has sufficient moment and torsional capacity to distribute the loading
over the full width of the sheet pile wall. Shear studs, hoops or brackets need to be
welded to the sheet pile and to steel plate girder deck beams in order to ensure
effective transfer of forces (see the Steel bearing piles guide(12).)
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
22
15 September 2003
The connection is formed by using a construction joint within the pile capping beam.
This enables the lower part to be cast around the High Modulus Piles and provides
a level surface on which to land the deck beams. The reinforcement cage in the
capping beam can then be completed and the concrete upper part cast.
The benefits of using a concrete pile cap beam may be summarised as:
C
Permitting the deck structure interface detail (e.g. holding-down bolts for a
steel superstructure or starter bars for a reinforced concrete superstructure) to
be attached to the heads of the piles.
As mentioned above, steel pile walls can be used with precast concrete beam decks an arrangement that has been used is shown in Figure 5.2.
High Modulus
Piles
Figure 5.2
Steel-to-steel connections
Where the bridge arrangements are such that spacing of deck beam and High
Modulus Pile UBs are similar, then individual steel connection details may be
considered. However, due to positional tolerances, potential fatigue problems, and
the on-site fit-up required, steel-to-steel connections of this type are considered
impractical at the present time.
23
15 September 2003
Steel bearing piles, in the form of box piles or tubular piles, are ideally suited to use
in this situation. As part of a frame abutment they will be subject to some induced
moment from the deck and the detail of this connection will be a very significant
aspect of the design. Moment is transmitted from deck girders to the column-piles
mainly in the plane of the girders, but transverse moments are also present. Careful
modelling of the column-pile stiffness is needed to prevent overdesign. It is very
easy to attract load to columns unnecessarily.
As for retaining wall abutments, a reinforced concrete pile cap is considered to be
the most practical means of achieving this connection, and a cross beam may be
needed to cope with the transverse moments. An elevational section of such a
connection is shown in Figure 5.3.
Asphaltic plug joint
Road construction
Deck beam
Construction joint
Tubular pile
Embankment fill
Figure 5.3
As for a frame abutment, a cross beam or capping beam would extend across the
width of the bridge deck. This therefore provides torsional restraint to the main
girders and transverse moment restraint to the columns when subject to lateral loads.
It can also act as an endscreen to retain the pavement formation. A recent example
is given in Figure 4.3, the Stockley Park canal bridge.
The design and construction of such a capping beam will be similar to that for a
frame abutment, but since the forces and moments will be less, the beam should be
less substantial.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
24
15 September 2003
5.2
An integral bankseat detail that is supported only by the soil is shown in Figure 5.4.
The steel beam ends are concreted directly into the bankseat endscreen wall.
Because the resistance to expansion provided by the soil behind an endscreen wall
is very low (due to its small depth) and an even more limited resistance to
contraction is provided by friction beneath a bankseat foundation footing, the
bankseat will therefore slide to and fro when the deck experiences significant
thermal strain. The bankseat will also rock as the ends of the beams rotate (about
a transverse axis) under live loading. In the design of such a bankseat the bearing
stress on the soil should therefore be kept at a lower level than for a static footing,
to allow for cyclic load degradation. Consequently the foundation width will be
larger; an inverted T-shape as shown will probably be preferable to an L-shape, for
stability.
Asphaltic plug joint
Deck beam
Select
granular fill
Movement by sliding
and rocking
Figure 5.4
The deck beams will have to be supported at a construction joint in the endwall and
torsionally restrained during construction until they are cast into the end wall.
25
15 September 2003
The pile cap must be sufficiently deep to permit a practical transfer of vertical load
from the pile.
Asphaltic plug joint
Deck beam
Select
granular fill
Reinforced concrete
bankseat capping beam
/endscreen wall
H pile
Figure 5.5
Piled bankseats can be used in combination with normal earthworks slopes in front
of the piles or with a retained face, such as provided by sheet pile wall or a
reinforced earth slope.
5.3
The provision of full moment continuity in frame abutment bridges necessitates the
construction of a complex connection between deck and supports, and requires
careful analysis of the bending interaction between the two structural elements. In
contrast, a more simple pinned integral design offers the benefits of integral
construction without the complexities of a moment connection.
The challenge in designing a pinned integral bridge will be to devise a pin that is
cheap to construct and durable over the full 120 year life of the bridge. The pin
bearing will need to accommodate relative rotation between deck and supporting
structure and transmit vertical and horizontal forces without relative displacement.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
26
15 September 2003
steel dowels (shear pins) in bearing plates between the capping beam and an
endscreen wall across the ends of the main beams. Such a detail is shown in
Figure 5.6.
Asphaltic plug joint
Reinforced concrete
end screen wall
Deck
beam
Reinforced concrete
capping beam
Sheet pile or
High Modulus Pile
Figure 5.6
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
27
15 September 2003
Figure 5.7 also shows an endscreen wall and the whole abutment is then a fallthrough arrangement similar to that used on the M25/A3 interchange at Wisley in
Surrey, only that was constructed in concrete.
Alternatively, a reinforced earth embankment face can be formed that will reduce
the total bridge span.
Asphaltic plug joint
Reinforced concrete
end screen wall
Deck
beam
Waterproof
membrane
Holding down bolts
set in grouted pockets
Studs for load transfer
Figure 5.7
5.4
A jointless deck type is created when a sliding or rocker bearing is inserted between
the bridge beams and the bankseat footing or pile cap as shown in Figures 5.8 and
5.9. Replacement of such a bearing is a little more difficult than in a conventional
bridge, because any jacking would raise the end wall as well and this will disturb
the adjacent road construction and surfacing. Cutting back and reconstruction of
that interface will therefore be required.
Asphaltic plug joint
Granular fill
Sliding
bearing
Figure 5.8
Flexible seal
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
28
15 September 2003
Granular fill
Sliding
bearing
Flexible seal
H piles
Figure 5.9
5.5
Intermediate supports
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
29
15 September 2003
Transverse
r.c. crosshead
beam
Tubular steel
column-piles
eg. 965 x 40 m
Figure 5.10 Two column-pile intermediate pier with r.c. crosshead beam
Either a cast in situ reinforced concrete transverse beam can be used (Figure 5.12)
or a pre-fabricated steel crosshead beam assembly (Figure 5.13). Steel crosshead
beams would be prefabricated, but a practical connection detail will have to be
developed for the connection to the pile. The detail shown has overlarge sleeves
that fit over the pile heads and once levelled either a welded connection is made or
the annulus can be epoxy grouted for load transfer.
Steel
crosshead
box beam
4 No. 762mm
x 40 mm w.t. steel
column-piles
Figure 5.11 Four column-pile intermediate pier with steel crosshead beam
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
30
15 September 2003
Where there are four column-piles, it may be possible to place an individual deck
beam directly onto its corresponding column-pile without a crosshead beam as
shown on Figure 5.11, but provision must be made for the tolerances that are
associated with driven piling (see Section 6.5). If no crosshead beam is provided
it will also be necessary to show that each individual column-pile can survive a
collision load case, where this is required, without losing its integrity. Where a
crosshead beam is provided, the collision load resistance is shared by the columnpile group.
Deck beam
Pinned bearing
Tubular pile
Figure 5.12 Tubular column-pile pier with pile capping beam and pinned
bearing
Moment connections between deck beams and intermediate supports should be
avoided as they offer little benefit but complicate construction and create potential
fatigue-prone details. Bearings should be used instead to transfer only horizontal
forces from the deck to the column-pile pier and to anchor the top of the column in
the event of vehicle collision resistance being required. A pinned bearing detail is
shown in Figure 5.12.
5.6
Bearings
31
15 September 2003
Rocker bearing
Knuckle bearing
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
32
15 September 2003
5.7
Skew bridges
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
33
15 September 2003
Where an integral bridge is supported on continuous walls, the behaviour of the end
walls is affected by this skew.
Consider a skew integral bridge on retaining walls (see Figure 5.15). The retaining
wall is able to flex only normal to its plane and, as a consequence, the bridge deck
will rotate in plan as it expands, as shown in the Figure (transverse expansion is not
shown - it does not affect the pattern of movement). The effect of high skew is to
magnify the displacement at the acute corner of a wide bridge, although this
magnification is only significant if the bridge is both highly skew and short in
relation to its width.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
34
15 September 2003
6.1
The type of continuous pile profile that is most appropriate for a retaining wall
abutment depends on the size and configuration of the bridge structure and the
magnitude of the acting axial and lateral loads.
For lightly loaded bridges of small span and/or retained height, where a pinned
connection is chosen for the deck to sub-structure connection, adequate abutments
may be formed from sheet piles. In the UK, two profiles, designated as Z and U,
are available; Z profiles are referred to as Frodingham sections, and U profiles as
Larssen sections. Although there are no real preferences for the use of Larssen
or Frodingham sections, each type of section does have its own characteristics,
which in certain situations can influence its choice.
For the larger bridge sizes and configurations where the structural capacity of U and
Z profiles is inadequate to accommodate the higher load magnitudes and/or retained
height, High Modulus Piles or interconnected Box Piles can be used.
Figure 6.1
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
35
15 September 2003
Figure 6.2
Larssen sections are stiffer and less prone to deviation from the intended placement
line than Frodingham sections. This is particularly the case when driving is carried
out through dense or difficult soil.
The range of Larssen steel sheet pile sections has also recently been increased to
include wider sections. These are named LX sections and have been introduced to
provide a superior strength-to-weight ratio pile that has comparable driving
capabilities.
The range of section modulus for LX profiles varies from 830 cm3/m for the LX8
section to 3201 cm3/m for the LX32 section, and for the Larssen profiles from
610 cm3/m for the 6W section to 5066 cm3/m for the 6 section. The complete range
of sections and their properties are listed in British Steels Piling Handbook.
Larssen 6 piles were used in the retaining walls for the frame abutment integral
bridge at the Bridgend-By-Pass, see Section 4.2.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
36
15 September 2003
11.7mm for 3N
10.4mm for 4N
11.9mm for 5
Frodingham section
steel sheet piling
x
11.7mm for 3N
14.0mm for 4N
17.0mm for 5
Universal beam
Figure 6.3
Currently British Steel offers High Modulus Piles formed from several profiles of
Frodingham section combined with a variety of Universal Beams, which are spaced
at 850 mm or 966 mm centres, depending on the section selected. Two Frodingham
sections (4N and 5) combined with eleven weights of Universal Beam are available
(see British Steels Piling Handbook).
High Modulus Piles are ideally suited to forming integral bridge abutments as they
have a suitable structural stiffness, and are practical to install. The universal beam
significantly increases the structural capacity of the section, and is capable of
supporting both lateral and axial loads. This type of section is most appropriate for
medium to large size bridges where the magnitude of loading and/or the retained
height rule out Larssen or Frodingham sheet piles. The spacings of the universal
beams (850 mm and 966 mm) are ideal for the construction of shallow bridge decks
formed of beams at close centres, but are sufficiently far apart to permit easy
inspection and maintenance.
High Modulus Piles were used in the A41 Stone Bridge replacement frame abutment
integral bridge at Aylesbury, see Section 4.2.
6.2
Box piles
Box piles are formed by welding two or more sheet pile sections together. Both
Larssen and Frodingham sheet piles can be used. They can be introduced into a line
of sheet piling at any point where local heavy loads are to be applied, for instance
beneath bridge beams (as in the Chad Brook Bridge, see Section 4.2), or used
separately. They are clutched together with adjacent sheet piles and can be
positioned in a sheet pile abutment so that its appearance is unaffected.
Frodingham plated box piles are formed by continuously welding a plate to a pair
of interlocked and intermittently welded sheet piles. Larssen box piles are formed
by welding together two sheet pile sections with continuous welds (see Figure 6.4).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
37
15 September 2003
Weld
Weld
Weld
Figure 6.4
Special box piles can be formed using certain combinations of sheet piles. Further
information can be obtained from British Steel Piling Technical Services.
6.3
Tubular piles
Tubular piles have been used as foundations for offshore steel frame structures for
over 60 years, ever since oil platforms were first required in the oil fields of Lake
Maracaibo in Venezuela in the 1920s. Initially, spare oilpipe was used for
convenience but, as the supporting structures became more sophisticated, the cold
rolling of piles in structural plate to project-specific diameters and wall thicknesses
has become more common.
Purpose rolled tubular piles are expensive but high quality steel linepipe that is
perfectly suitable for piling is available throughout the EU at reasonable cost.
British Steel manufacture such pipe at their new Hartlepool linepipe mill and can
fabricate project-specific sizes at their Stockton-on-Tees mill. See Section 7.2.4 for
linepipe material specifications.
Linepipe is manufactured to a different product specification to structural steel, but
its properties are suitable for most structural applications as well. The cold-rolling
process produces consistently higher yield strengths than those for hot-rolled steel
products and this can have significant benefits for highly loaded bearing pile and
structural column-pile applications, and can permit harder driving. Selection criteria
are often based on the accommodation of high driving stresses during installation
and the resistance to lateral loading shear forces in service without inducing plastic
deformation in the section.
An example of a project where tubular bearing piles were used is the Stockley Park
canal bridge, see Section 4.2.
The range of linepipe sizes produced by British Steel in 1996 is shown in
Figure 6.5.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
38
15 September 2003
Wall thickness
(mm) (ins)
50
2.0
45
1.8
40
1.6
35
1.4
30
1.2
25
1.0
20
0.8
15
0.6
10
0.4
0.2
API 5L X52
API 5L X65
API 5L X70
API 5L X80
16 18 20 22 24 26 28 30 32 34 36 38 40 42
Outside pipe diameter (ins)
Figure 6.5
6.4
H-Piles
Steel H-piles have been used extensively in bearing pile supports for bridges in the
USA and the UK. A recent example is their use on the A47 Walpole to Tilney End
By-Pass bridges in Norfolk, designed by Stirling Maynard & Partners.
Steel H-piles (see Figure 6.6) are very efficient in surface area provision for shaft
friction piling. For example a 305 mm 305 mm x 110 kg/m H-pile section has
an external surface equivalent to a concrete pile of diameter 601 mm and a
displacement volume only 5% of that of the concrete pile, which enables it to be
driven with less energy and into more dense soils. The displacement volumes of
British Steels range of 305 mm 305 mm piles cover a range of 3% to 8% of that
of equivalent concrete piles. In any given foundation plan area therefore, a greater
number of steel piles can be provided in a group with a standard spacing of 3B (or
3Dia.) than concrete piles and the load supported can either be greater, or if soil
conditions permit, the driven steel piles can be shorter to support a given structural
load. Further information is given in the Steel bearing piles guide(12).
y
Figure 6.6
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
39
15 September 2003
H-piles are also used very effectively to transfer bearing loads down to a buried
rockhead that may underlie the site and to get round buried obstructions. It is often
advisable to use special pile shoes to strengthen the tip and prevent damage or
buckling under hard driving conditions. Further information is given in the Steel
bearing piles guide(12).
6.5
Installation tolerances
25 to 75 mm
Dependant on equipment used.
2 mm
of pile toe
120 mm
1.5%
1%
1%
0.5%
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
40
15 September 2003
damaged during installation. Where dramatic changes in toe level are anticipated,
e.g. a geologically eroded buried rock surface, steel piling is particularly appropriate
but the advice of specialist piling contractors should be sought during conceptual
design of the bridge foundation. The Federation of Piling Specialists should be
consulted on such a matter. Sometimes, driving causes previously driven piles to
rise but this can be compensated by redriving these piles.
It is inconceivable that an acceptable level can be achieved at the top of the pile to
avoid any need for on-site cutting. However, precision mechanical, burning and
hydraulic grit blast cutters are available to achieve a level cut within 1 to 2 mm
accuracy.
6.6
Increasing attention has been directed to the environmental effects of pile driving in
recent years. Although the duration of the piling contract may be short in
comparison with the whole contract period, noise and vibration may be perceived
more acute during the piling phase. This is exacerbated because pile driving is often
the first construction activity on site. Forewarning of the public is therefore always
a worthwhile precaution before starting pile driving operations.
In the UK, the Control of Pollution Act (1974)(31) provides a legislative framework
for, amongst other things, the control of construction site noise. The Act defines
noise as including vibration and it provides for the publication and approval of
Codes of Practice, the approved code being BS 5228(32). A section of the Code (Part
4) deals specifically with piling noise. This Code was revised in 1992 to include
guidance on vibration. Detailed information is given in the Design guide for steel
sheet piled bridge abutments(30).
The study and codification of noise and vibration as a necessary part of construction
work has helped with its formal acceptance and treatment. The construction industry
has developed new driving equipment and methods to reduce the levels of noise and
vibration during pile driving which help to keep them within acceptable limits.
Driven piles are hence acceptable from the environmental aspect on many more sites
than they used to be.
6.7
Driveability
The designer needs to ensure that the steel piles that he has chosen for his concept
can be installed to the desired penetration.
Several simple driving formulae have been used by piling specialists over many
decades and reliable empirical rules derived from this experience to use in those
formulae. Two of the most reliable formulae are those of Hiley and Janbu and these
are explained in detail in Cornfield(33) and in the Steel bearing piles guide(12).
Experienced judgement is required on hammer efficiency and on driving resistance
in order to obtain sensible results. A nomogram is included in the SCI publication
to assist the user. If the designer has little experience of assessing driveability,
assistance can be sought from a specialist piling company recommended by the FPS.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
41
15 September 2003
The driving resistances of sheet piles and individual bearing piles differs
fundamentally because of the interclutch friction with sheet piles. In fact, this has
been found to be the most significant resistance with sheet piles and explains why
single bearing piles can be driven to much greater depths than sheet piles. The FPS
and British Steel have published Tables of maximum driveable depths for sheet piles
in different types of soil for the guidance of designers. These are reproduced in the
Steel bearing piles guide(12) and Design guide for steel sheet piled abutments(30).
A more sophisticated method of assessing pile driveability for bearing piles is to use
a computer version of the wave equation. The WEAP program is the most highly
developed and the most extensively used in design offices. It contains a complete
hammer data library and guidance on soil resistance and reliable default values for
all the parameters involved that have been derived from the back analysis of
instrumented pile driving.
6.8
Corrosion allowances
The means for countering the effect of corrosion of steel piles are well developed.
Guidance is presented in British Steels Piling Handbook and in the Highways
Agencys BD 42 Design of embedded retaining walls and bridge abutments(35).
BS 8002(36) and BS 6349(37) both consider that the end of effective life of a steel sheet
piles occurs when any part of the pile reaches its maximum permissible stress as a
result of the reduction in section due to corrosion. A pile section chosen for the
in-service condition should therefore also be adequate at the end-of-design-life (i.e.
after 120 years, when designed to BS 5400), at which time the effective pile section
will have been reduced by corrosion, but also the forces in the pile may have been
modified by the change in stiffness of the pile. It is not immediately obvious
whether the start of in-service life condition or the end-of-design-life condition will
be critical for the structural design of the abutment, it depends on the actual
situation.
As the corrosion loss allowance varies along the pile according to the corrosion
environment, the designer needs to be aware that the maximum corrosion may not
occur at the same level as the maximum forces and moments, and to allow for this
accordingly.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
42
15 September 2003
The design procedure for use of sheet piled walls in integral bridges is illustrated in
the Integral steel bridges: Design of a single-span bridge - Worked example(5). In
that example it is found that the redistribution of earth pressure that occurs as a
result of the increased flexure of a corroded section is significant, and that the endof-life condition is a critical design load case in the selection of pile section.
Exposure zone
Sacrificial thickness
(for one side of the pile only)
(mm)
Atmospheric
The use of a higher strength steel than required if no corrosion were assumed
(i.e. use steel grade S355GP, to BS EN 10248(38), in a wall designed for steel
grade S270GP). This permits a greater loss of metal before stresses become
critical.
Concrete encasement of steel piles above the water line. However, the inherent
impact absorbing properties of the pile are lost by this method.
Details of these options are given in British Steels Piling Handbook.
However, where the exposed face of a retaining wall is potentially exposed to salty
road spray (i.e. a splash zone) it will probably be more effective and aesthetically
pleasing to use a non-structural cladding or a protective coating, rather than a large
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
43
15 September 2003
The nature of in situ fill soils can be variable and a full soil analysis will be required
to assess the likely corrosion performance of steel in the environment. Soil tests to
determine the pH of the soil should be in accordance with BS 1377: Part 3(41) and
as directed by the Contract to determine resistivity. Other tests may be relevant,
and most of these are reviewed in CIRIAs series of reports on contaminated land
(contact CIRIA for further details).
In a controlled fill (i.e. selected granular fill, as referred to in Clause 3.8 of
BD 42), no special measures are required and the same corrosion rates as in natural
undisturbed soils can be assumed.
Further advice on corrosion assessment and protection can be obtained from British
Steel, Swinden Technology Centre, or from SCI.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
44
15 September 2003
Exposed to
atmosphere ?
Yes
No
Completely and
continuously
immersed in water?
Yes
Aggressive
atmosphere /
water / effluent ?
No
No
Sacrificial thickness 4 mm
Yes
Seek specialist advice
Splash
zone ?
Yes
Cladding ?
No
No
Sacrificial thickness 9 mm
Yes
Sacrificial thickness 4 mm
Natural Undisturbed
soil ?
Yes
Sacrificial thickness 2 mm
No
Classify soil from Tables 1, 2
and 3 of BD 42/94
Non-aggressive soil ?
Yes
Sacrificial thickness 2 mm
No
Very
aggressive soil ?
No
Sacrificial thickness 4 mm
Yes
Seek specialist advice
Figure 6.7
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
45
15 September 2003
7 DESIGN BASIS
7.1
General principles
Limit state design should be adopted as far as possible for integral and jointless deck
bridges. Where only rules on an allowable stress basis are available, they must be
applied in a way that is consistent with the limit state design used elsewhere.
Compatibility must be maintained between the displacements and deformations of
the deck structure, the foundations and the connections between them. An overall
analysis is required that takes account of the interaction between structural and
geotechnical elements (such an analysis is not yet covered in any design code).
7.2
Design standards
For bridges the applicable National Standard is BS 5400(42). This covers the design
of steel, concrete and composite structural elements of bridges. Composite bridge
deck design is normally in accordance with Parts 3, 4 and 5 of BS 5400(42).
For the design of earth retaining structures and foundations, reference may be
made to BS 8002(36) and BS 8004(43) .
BS 8002
BS 8002 is primarily applicable to small and medium sized earth-retaining structures
with a retained height of up to approximately 8 m, although it is stated that many
of the recommendations are more generally applicable. BS 8002 is a code of
practice based on the use of a simplistic and traditional limit equilibrium method for
retaining wall design that has been used to ensure stability of walls. This method
is based on the use of theoretical limiting earth pressures without practical proof that
they can co-exist at the point of wall overturning. This method has not been related
to wall movement nor have trial walls demonstrated these earth pressure profiles for
all types of wall, and therefore the method cannot be used reliably for analysis of
retaining walls subject to displacement.
As BS 8002 is based on this simplistic approach, it cannot uniquely define the limit
states that are to be used for design. This is seen in its definition for limit state
design in that the safety and stability of the retaining wall may be achieved,
whether by overall factors of safety, or partial factors of safety, or by other
measures. Owing to this lack of precision in definition, BS 8002 refers both to
partial factor based limit state codes of practices such as BS 8110, BS 5400 and
BS 5950; and to BS 449: Part 2(44), which is a working stress code of practice
based on the use of a lumped factor approach.
BS 8002 uses an approach based on worst credible soil and ground parameters to
develop an adequate margin of safety(45). A mobilised soil strength is advocated
for use in design at the serviceability limit state only (to limit retaining wall
movements), because BS 8002 states that the most severe earth pressures that can
credibly occur on a retaining wall, occur at that limit state. The mobilised soil
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
46
15 September 2003
Owing to their more complex behaviour, integral bridges require more precise
analysis methods to be used than the limit equilibrium wall stability methods.
BS 8002(36) is limited in its application to wall stability type analysis; it does not
cover the soil-structure interaction that is required in order to determine bending
moments, forces and stresses on integral bridge retaining walls. For highway
structures, HA Departmental Standard BA 42/96(9) advises that BS 8002 should be
used, but only for particular aspects of design. It does require, however, that steel
embedded retaining walls for integral bridges are designed in accordance with
BS 5400 Parts 3, 4 and 5 using the limit state partial factor approach and not the
working stress approach.
BS 8004
BS 8004 is applicable to the design and construction of foundations in general,
which can be piled or shallow bearing foundations. BS 8004 is based on a working
stress approach using lumped factors of safety and uses a design approach for
foundations based on moderately conservative soil parameters, loads and geometry,
and on generous factors of safety.
Eurocodes
Recently two CEN documents have become available. Eurocode 7 Geotechnical
Design(48), issued by BSI as DD-ENV 1997-1, and the draft prestandard Eurocode 3:
Part 5 Design of Steel Structures - Piling(25) has been circulated in industry as
prENV 1993-5(25). These are fundamentally more rigorous documents which apply
limit state principles and use a partial factor approach.
Use of Eurocode 7(48) can be of some help in the design of integral bridges since it
uses limit state design with partial factors that are compatible with the design
approach of BS 5400. The use of Eurocode 7 is endorsed by the Highways Agency
in BA 42/96.
In Eurocode 7, the partial factors which are applied to the characteristic value of the
soil parameter are presented in Clause 2.4.2, Table 2.1(48) and that Table is
reproduced below for information. Annex B of DD ENV 1997-1 provides additional
recommendations relating to these partial factors.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
47
15 September 2003
Table 7.1
Case
cu
qu
1.50
1.1
1.3
1.2
1.2
1.00
1.50
1.0
1.0
1.0
1.0
1.00
1.30
1.25
1.6
1.4
1.4
Un-favourable
Favourable
Un-favourable
Case A
1.00
0.95
Case B
1.35
Case C
1.00
Notes:
Case A
Case B
Case C
N
cu
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
48
15 September 2003
X46
317
X52
358
X56
386
X60
413
X65
448
X70
482
X80
551
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
49
15 September 2003
7.3
Limit state design, as set out in BS 5400, requires that the design resistance R* is at
least equal to the design load effects S*. The partial factors to be used are (fL on
loads, (m on material strengths and (f3 to take account of inaccurate assessment, etc.
The partial factors are applied in two different ways, depending on the Part of
BS 5400 concerned. In Part 3 the design values are expressed as:
R* =
function(characteristic strength)
f3 m
*
and S = ( effects of
fL
design loads)
R* =
function(characteristic strength)
and
S* =
f3
( effects of
fL
design loads )
Consequently, care has to be exercised in the application of the partial factor (f3
when dealing with a mixture of steel and concrete elements. This situation also
arises in the design of steel retaining structures, since BD 42 also requires that (f3
is applied to the load effects; when that is the case, the factor should be omitted
from the calculation of design resistance, even for steel elements.
Design resistances are determined in accordance with BS 5400: Parts 3, 4 and 5, for
the steel, concrete and composite elements respectively. The integral bridge deck
and High Modulus Piles are designed to BS 5400: Parts 3 and 5. The capping beam
is designed to BS 5400: Parts 4 and 5.
Design of sheet piling to date has used the elastic section properties of the wall, but
there is a move towards using plastic section properties for the ultimate limit state
design of sheet pile walls. This is detailed in the new Eurocode 3: Part 5(25).
Development of fully plastic structural capacity of U-section sheet piles (Larssens)
is only obtained if pairs of sections are crimped or welded at the interlocks.
Z profiles such as Frodingham sections do not suffer from this problem. Guidance
to determine the appropriate capacity of U sections is given in British Steels Piling
Handbook(24).
In drivability assessments, the calculated dynamic stresses are compared with 0.9 fy
for the steel grade being used.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
50
15 September 2003
7.4
Design load effects on the foundations of an integral bridge arise from a variety of
sources:
C
Soil pressures.
Seepage forces.
Ground movements.
Traffic loads.
Braking loads.
Movements at the ends of the bridge deck due to change of the effective
temperatures of the bridge are partially restrained by the resistance provided by the
adjacent soil, leading to forces and moments in the bridge deck and foundations.
Reference should be made to BA 42(9) for the specification of thermal strains and the
determination of movements and restraint forces in the bridge deck and foundations.
7.5
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
51
15 September 2003
A monitoring plan is set up which shows that the actual behaviour lies within
the acceptable limits.
For small and medium sized structures, the wall displacements will be small and the
inherent uncertainties are normally catered for in design by adopting conservative
values of soil properties (see Section 10). For larger and more complex structures,
however, any over-conservatism may lead to unacceptably high costs.
7.6
Design report
Stability analyses of the site and calculations for the structural design of the
steel sheet pile wall.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
52
15 September 2003
7.7
Cyclic loads that can cause fatigue damage to integral bridges are those due to
vehicular movement of traffic along the bridge deck and those due to expansion and
contractions of the bridge deck due to temperature fluctuations. For both welded
and bolted steel structures, fatigue life is normally governed by the behaviour of the
connections, which include main and secondary connections. The optimal fatigue
behaviour is obtained by ensuring that the structure is so detailed and constructed
that stress concentrations are kept to a minimum and where possible the elements are
able to deform in their intended ways without introducing secondary deformations
and stresses due to local restraints.
In framed integral bridges, the moment connections at the ends will need to be
checked for fatigue due to traffic loads. Bolted and welded details, and shear
connectors, will need to be considered.
The question also arises of whether variation in temperature, leading to strains and
movements, might also give rise to fatigue.
To assess the extent of fatigue that could be present, it is necessary that reliable data
pertaining to temperature variation be used. Currently, limited temperature data is
published and four possible sources are those given in BS 5400, BS 7608(61),
TRL 696(62) and TRL 765(63).
However, for a general appreciation, reference can be made to data in a paper by
Hambly and Owens(64) in which the question of fatigue assessment of thermally
restrained bridges has been considered. In this case the bridge considered was a
steel box girder. BS 5400, TRL 696 and TRL 765 were used to provide data and
included three distinct types of temperature cycles:
C
One extreme cycle during the 120 year design life between the extreme
maximum and minimum temperatures (from BS 5400).
120 annual cycles between summer and winter temperatures (estimated from
TRL 696).
365 120 daily cycles between day and night temperatures (estimated from
TRL 696).
Time Period
Cycles
Temperature changes
Effective
EC
Difference
EC
70 + 20
90
120
50 + 20
70
44,000
14 + 14
28
Clearly, the number of cycles involved is much lower than that normally considered
for design (BS 5400: Part 10 gives a limiting stress range of over 200 N/mm2 at
105 cycles, the lowest number of cycles normally considered). Fatigue due to
thermal variation is therefore extremely unlikely in structural elements.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
53
15 September 2003
8 DESIGN METHODOLOGY
8.1
Design sequence
For a non-integral bridge, the design of the deck structure and the design of the
foundations are essentially separate exercises. The main interaction is simply a
transfer of support reactions from the deck and the detailing of the bearing areas.
For an integral bridge there is a degree of continuity that creates interaction between
the design of the two elements. The stiffness of the soil reactions and of the
connection will affect the stresses induced and the construction sequence may affect
the loading combinations that will apply. Therefore a decision must be made on the
type of foundation that is to be used at the concept and scheme costing stage.
For a jointless deck type bridge where only the deck is integral, the old design
procedures of independent deck and substructure design may be used, since the
elements are separated by bearings that allow relative movement.
The overall design sequence for an integral bridge is illustrated in Figure 8.1. As
can be seen, the first four activities concern the conceptual design of the bridge, and
within this phase it is necessary to decide the configuration that is the most
economic, taking into account the site constraints, the practicalities of construction
and whole life costing. This must be done before the stage of the global interaction
analysis is reached, because the latter requires realistic structural element sizes and
stiffnesses.
Some experience of soil-structure interaction analysis and familiarity with the effect
of different soil types on wall and pile behaviour will greatly assist in judgements
on the relative merits of different wall constructions, on the effect of different
connections or of the merits of piled versus spread footings in their response to the
deck thermal movements and braking loads.
A well structured methodology for analysis and design is important because it
enables the designer to focus not only on the immediate activity at hand but also on
what is to be achieved in the overall design. Guidance given today tends to be less
prescriptive than in the past because of the need to allow the application of new
knowledge and the development of specialised computer software. This approach
allows the designer to think, rather than just rigidly apply a set of prescribed
instructions, and is particularly valuable in the analysis of integral bridges, for which
formalised rules have not yet been established.
The overall design sequence is expanded into a more detailed methodology for
embedded retaining wall type bridge abutments in Section 8.3, and for piled column
solutions for fall-through abutments and intermediate pier supports in Section 8.4.
These methodologies are based on the currently available Standards identified in
Section 7, supplemented by recently developed practice that employs numerical
analysis software to solve the soil-structure interaction problem (as discussed in
Section 8.3.7). Each aspect of analysis and design is covered, together with a
sequence in which the activities should be undertaken in order to design retaining
wall abutments, and tubular steel column-piles most effectively. In Sections 8.3 and
8.4 the reader is referred to the appropriate sections of this document that explain
the detailed considerations and to other references.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
54
15 September 2003
Site investigation
Design deck
Evaluate benefits of
continuity deck/support
8.2.1
8.2.2
8.2.3
Design abutment or
bankseat structure
Intermediate
supports?
Yes
No
Include continuity in
interaction
Carry out interaction
analysis
(deck + supports)
Check adequacy of
deck and supports
Strengths OK
?
No
Yes
Design details,
including connections
deck/supports
Figure 8.1
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
55
15 September 2003
8.2
Preliminary stages
56
15 September 2003
expansion will add to these stresses. If the bottom flanges are not adequately
restrained laterally by the cross bracing, the effects of buckling in hogging regions
near the end supports may need to be analysed.
For an analysis of a frame integral bridge, in the first instance it should be assumed
that fully fixed conditions apply at the bridge deck supports. Design load effects
(moments and shears) in the deck can then be determined. An initial estimate of
axial load should be made to allow determination of axial stresses. Flange and web
sizes can then be verified before proceeding to detailed design. Moments and shears
should also be determined assuming pinned supports, so that any benefits of moment
continuity may be evaluated.
For a bridge that is presumed from the outset to be a pinned integral bridge, a
similar exercise should be carried out, except that the fixed supports condition is not
considered.
No detailed advice is given in this publication on bridge deck design; guidance is
available from other SCI publications(65).
8.3
Figure 8.2 below shows a design sequence for the design of retaining wall
abutments. Both frame abutment and pinned integral bridges are covered. The
activities in the preliminary stage outline above are common to both types.
The following Sections deal with each of the activities shown in Figure 8.2.
The design methodology concentrates primarily on the use of numerical analysis
methods for solving soil-structure interaction.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
57
15 September 2003
8.3.2
8.3.1
8.3.3
8.3.4
Frame abutment
Pinned abutment
Figure 8.2
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
58
15 September 2003
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
59
15 September 2003
To model the soil/structure interaction between the deck, substructure and the
adjacent soil, compatible boundary conditions at the connection need to be
established. The modelling of the interaction is dependent on the types of computer
program that are used to analyse (separately) the deck structure and the substructure.
Modelling is discussed in Section 11.
Frame abutment
A frame abutment has both rotational and displacement continuity between deck and
supports.
Typically, a rotational spring will need to be applied to the top of the wall to
represent the effect of deck stiffness on the pile capping beam, and another spring
will need to be applied to the deck model supports to represent the restraint of the
wall.
The spring representing the deck can be calculated from the deck analysis model
with an end moment applied, because the M/N relationship is linear elastic. When
assessing the M/N relationship, care should be taken to ensure that the deck response
corresponds to the loadcase under consideration ( e.g. longitudinal load produces
a sway response).
Similarly, the axial stiffness of the deck may be easily calculated. The wall analysis
model may include this as a spring, although the propping effect of most decks can
be modelled as an infinitely stiff prop.
The soil-response is potentially non-linear, therefore the M/N characteristics of the
head of the abutment wall should be evaluated over a range of moment values up to
the full fixed end moment from the deck structure.
Pinned-integral
When using numerical analysis models the same aspects apply as for the frame
abutment situation but without the moment continuity, so the rotational spring is not
required.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
60
15 September 2003
because the designer must consider long and short term deck stiffness(due to
concrete creep effects) together with uncorroded or corroded (end of life) steel pile
wall thickness.
Pinned-integral
Numerical analysis programs like FREW and WALLAP can be used to determine
the bending moment and force diagrams for pinned integral bridges as well, but a
more simple limit equilibrium based propped cantilever analysis like ReWaRD is
also generally sufficient for low to medium height walls (less than about 8 m in
height).
For low to medium height walls it has been found that limit equilibrium analysis can
give comparable bending moments in the sheet pile wall to those given by FREW
and WALLAP, provided that the wall section is relatively stiff and most of the
bending induced is due to rotation of the wall. In high walls with greater
slenderness, the ability of steel sheet pile walls to deflect, and thereby to redistribute
earth pressure, can only be satisfactorily analysed using the more sophisticated
numerical analysis programs like FREW and WALLAP.
In view of the lack of reality in limit equilibrium calculations, it is prudent to
determine the maximum bending moment using several methods, as recommended
in CIRIA 104(66) and BD 42(35), and to weigh the effects of small increases in depth
of embedment; variations in soil parameters; and of possible loading conditions
before selecting the design value.
The application of the deck thermal expansion and contraction movements will give
a problem when using limit equilibrium methods, since the models are based on
applied earth pressure forces, not strains. It is not advised to use limit equilibrium
methods for bridge decks longer than 10 m, where displacements become
appreciable and can be expected to affect earth pressures significantly.
Corrosion effects
The effects of corrosion need to be considered when obtaining the total bending
moments and forces acting on the substructure:
The pile section properties and stiffness reduce with decreasing wall thickness due to
corrosion, and therefore bending moment and force magnitudes acting on the wall will
change over the service life time of the bridge. Separate analyses are required to
cover this, one for the uncorroded state and the other for the fully corroded state.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
61
15 September 2003
The design of the reinforced concrete capping beam involves considering the force
and moment transfer between the cap and the embedded pile tops; the cap and the
end of the embedded deck beam; and the transverse load and moment transfer across
the wall.
Reinforcement has to be provided to effect all these three load transfer mechanisms
for frame integral abutments. This is illustrated in Steel integral bridges: Design of
a single-span bridge - Worked example(5). A section through that detail is shown in
Figure 8.3.
Construction Joints
Neutral Axis of
High Modulus Pile
Figure 8.3
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
62
15 September 2003
Reinforcement detailing for a typical pile capping beam for a pinned-integral bridge
is covered in Steel integral bridges: Design of a multi-span bridge - Worked
example(3).
A construction joint will need to be provided at a convenient level in the capping
beam that is above the head of the embedded pile on which to land the deck beams.
A strip of elastomeric material can be used as a soffit to take up irregularities in the
concrete surface or steel wedges or packing plates used to level the beams as
necessary.
An illustration of load transfer shear studs in the top of a tubular column-pile is
given in Figure 5.7 and Figure 5.12.
8.4
Figure 8.4 below shows a design sequence for the activities that are specific to a
steel tubular column-pile design for intermediate piers and fall-through abutments.
Reference to specific design guidance, both in this document and in other
publications on the design of column-pile abutments and intermediate supports, is
given in the same manner as the preceding Section 8.3 does for embedded walls.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
63
15 September 2003
8.4.1
8.4.2
Pinned integral
8.4.3
8.4.3
8.4.3
Check deck
Figure 8.4
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
64
15 September 2003
This first requires output from the deck analysis that gives the order of rotation at
the soffit of the pile capping beam, assuming different restraints from the pile.
Since there may be only 4 to 6 tubular steel piles across a typical overbridge
endscreen wall, the first assumption should be no rotational restraint, only lateral
restraint i.e a pinned-support case for the deck.
This rotation is then applied to a fixed head pile model, like ALP, and the lateral
analysis is run to derive the moment and shear force at pile head versus pile rotation
relationship and the combined stress in the pile section. The pile length may need
to be extended to achieve toe fixity in this analysis.
If the induced reaction force from the pile at the head is significant, an appropriate
force per unit length is then applied to the capping beam to produce a beam moment
diagram, to permit design of the connection with the pile and the bending across the
beam to add to that from earth pressure on any endscreen wall. The deck analysis
may need to be re-run to revise the moment diagram if the pile restraint is sufficient
to cause a hogging moment at the deck/capping beam level.
Pinned integral bridge
Unlike abutment walls, there is no complication for wall stability analysis, so the
basic design work for using column-piles in a pinned integral bridge is completed
by checking the combined stresses from the axial and lateral loading analysis.
In respect of end of embankment stability, there may also be a deep-seated slip plane
that passes across the piles at some depth and the shear force from that will have to
be applied to the pile as well.
Pile driving stresses should be checked for the final section.
8.5
The steel plate girder beams of a composite bridge deck will have to be cast into the
reinforced concrete endscreen wall of a bankseat substructure in order to create an
integral bridge. The geometry and the degree of moment continuity will dictate the
design of the connection.
Where the foundation soils are good and construction depth is less than about 2 to
3 metres, it will be possible to use a spread foundation, otherwise piled foundations
may be the most economic solution (as used on the A1 North Shotton overbridge).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
65
15 September 2003
The design of endscreen walls for lateral earth pressures is covered in the
recommendations of BA 42/96(9) by the Highways Agency which also recommends
a reduction in the allowable bearing pressure beneath spread footings to allow for
the effects of cyclic loading. Otherwise conventional design of the substructure and
foundations to BS 5400(42) and BS 8004(43) is used, as appropriate.
Where bearing piled bankseats are required, the design can be performed in a
similar manner to that described in Section 8.4.
8.6
Deck design
For comment on the effects to be considered in deck design of integral bridges, see
Section 15.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
66
15 September 2003
9.1
Standard Penetration Test resistance N values for pile end bearing and shaft
resistance.
Deformation and stiffness soil properties required include but are not limited to:
C
Poissons ratio.
Consolidation characteristics.
Typical values for appropriate soil properties are presented in the Design guide for
steel sheet piled bridge abutments(30).
9.2
Site investigation
Site investigations should be carried out in accordance with BS 5930(69) and methods
of in situ and laboratory testing to BS 1377(41). Advice is also given by the
Highways Agency in Advice Note HA 34/87(70). A series of informative CIRIA
reports(71)(72)(73) is also available advising on site investigations.
Accurately determined soil properties, both local and adjacent to the construction
site, need to be obtained to enable a bridge abutment or pier to be designed
accurately and confidently. Soil parameters must not be determined in isolation and
need to be presented with information relating to the physical conditions in the
vicinity of the structure. This includes the topography of the site, details of adjacent
foundations and services, and the nature of the ground water conditions.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
67
15 September 2003
9.3
The determination of soil parameters for design needs to be based on the careful
assessment of a range of values of each parameter that might govern the
performance of the structure during its design life, with account taken of the
conditions representative of the ground and the nature of the surrounding
environment.
The assessment of appropriate parameters is often dependent on the mechanism or
mode of deformation being considered. Strain levels and compatibility should
therefore be considered in the assessment of strengths in materials through which a
presumed failure surface can occur. Ranges of values may also be required,
particularly if the soil parameter values are likely to change during the lifetime of
the embedded sheet pile abutment structure. Typically for structures in clays it is
necessary to obtain soil parameters both for short-term and long-term conditions.
This requires that soil parameters for both drained and/or undrained conditions are
obtained from soil tests.
9.4
The uncertainty involved in the selection of soil strength parameters can be catered
for by specifying that soil properties are based on the definitions stated in
BS 8002(36). BS 8002 requires that representative values are obtained, where a
representative value is defined as a conservative estimate of the property of the soil
as it exists in situ. In this case conservative is defined as a value of soil
parameter which is more adverse than the most likely value. It may be less or
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
68
15 September 2003
greater than the most likely values and tends towards the limit of the credible range
of values. It is comparable with the definition of the worst credible value referred
to the CIRIA Report Design of retaining walls in stiff clays(66).
Shear stress
100
Moderately conservative
o
(c'=10 kN/m 2 , ' = 21 )
50
Worst credible
o
(c'=0 , ' = 21 )
50
100
150
200
Figure 9.1
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
69
15 September 2003
10
This section presents the method that is used to assess the overall stability of
embedded retaining walls against overturning and determine a depth of embedment
such that horizontal wall displacements are within serviceability requirements.
Other possible modes of failures may occur but they are less likely than that of
overturning. These include stability failure by deep-seated slip planes passing
behind and below the wall, which can be particularly important for waterfront
structures or in sloping ground.
Embedded retaining walls rely on the resistance of the ground below excavation
level and, where present, on the forces provided by anchors for their stability. For
both support conditions, whether for cantilever or propped walls, it is assumed that
there is enough movement of the ground to allow full active and passive pressures
to be generated at limiting conditions.
The classical methods of determining earth pressure for the limit equilibrium wall
stability analysis are the same as those used for non-integral bridges. For further
explanation see the Design guide for steel sheet piled bridge abutments(30). Simple
hand calculations can be used, but computer programs like British Steels ReWaRD
program can save time and permit more rigorous comparison of several methods in
the judgement of embedment depth.
The limit equilibrium approach is simplistic in that it does not satisfy all the
fundamental theoretical requirements to simulate soil-structure behaviour. No
account is taken of the mode of wall displacement on the resultant earth pressure and
no indication of the distribution or magnitude of earth pressure prior to ultimate
failure is given. The estimation of structural forces in the embedded retaining wall
under serviceability conditions is therefore extremely difficult. Although a great
deal of experience has been gained using this approximate method, the above
shortcomings limit its use to initial wall stability analysis and determining the depth
of embedment of retaining walls. Bending moments and forces acting on the wall
can only be determined accurately if soil-structure interaction using numerical
analysis techniques is considered (see Section 11).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
70
15 September 2003
The former can be allowed for either by using an unfactored worst credible
parameter, or by using a factored moderately conservative parameter, or the
representative/1.2 as in BS 8002(36).
Since weaker soils produce larger structural displacements at the point of limit
equilibrium, in-service displacements can be limited by the application of a further
factor. This factor would be applied to either the unfactored worst credible
parameter or to the factored moderately conservative parameter.
In practice both functions are grouped together in one lumped factor. This practice
is at odds with the trend towards discrete partial factors in structural engineering for
three main reasons.
C
Soil parameters are highly variable both within one site and from site to site.
They do not lend themselves to statistical distributions that lie at the heart of the
partial safety factor approach.
Limit equilibrium methods are widely used, hence there is a need for a factor
to limit deflections.
Insufficient test data for calibration from actual walls is available to assess the
accuracy of the deflection predictions of existing numerical-based software (i.e.
FREW, WALLAP).
Until the above circumstances change, there will still be a need to carry out a limit
equilibrium analysis to establish an embedment depth that sufficiently limits
serviceability limit deflections.
As a result of the recent introduction in the United Kingdom of BS 8002 and the
draft for development version of Eurocode 7(48), the Factor on Strength method is
increasingly being used for stability determination. Also the Burland and Potts(74)
method is increasingly being used because it is more consistent than, and free from
the peculiarities contained in, the other lumped factor methods (see Section 10.2.2).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
71
15 September 2003
Discuss me ...
10.2.1
In this approach the soil strength parameters used to derive the earth pressure
coefficients are reduced by dividing by appropriate factors. The divisor can be a
factor which represents the mobilised soil strength acting at the serviceability limit
state (i.e. in accordance with BS 8002) or partial factors applied in an ultimate limit
state (i.e. in accordance with Eurocode 7(48)).
In an effective stress analysis the soil strength parameters include c', the effective
cohesion and N' the effective angle of shearing resistance. To allow for the
uncertainties associated with c' and N, two factors of safety are defined, Fc and FN.
The factored parameters for effective stress analysis are termed mobilised values and
are given by:
c m'=
c'
tan '
and m'= tan 1
Fc
F
where cm' and Nm' are the mobilised values of the respective strength parameters
c' and N'.
To maintain overall consistency
m
cmw ' cw
=
and
=
m ' '
cm '
c'
where *m and cwm are the mobilised values of wall friction * and wall adhesion cw.
For total stress analysis:
cum =
cu
Fc
where cum is the mobilised value of the undrained shear strength cu.
It is common when performing an effective stress analysis to assume that Fc = FN
= Fs, although in Eurocode 7(48) individual factors for c' and N' are given.
The mobilised strengths are used to calculate the earth pressure coefficients and the
distribution of earth pressure on the wall. The factored strength parameters increase
active and reduce passive earth pressures and modify the relative distribution of
these pressures.
The resultant forces on the back and front of the wall are then expressed as a
function of the unknown depth of embedment, d. By equating moments to zero the
embedment depth can be calculated.
The design embedment is that for which the following relationship is satisfied:
M fp = M fa + M w
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
72
15 September 2003
where,
Mfp =
Mfa =
Mw =
When the forces are expressed as functions of embedment depth do, this equation
reduces to a cubic expression in terms of do.
The factor to be used in the Factor on Strength method is given in Table 10.1.
Table 10.1 Soil strength factor Fs
Reference
Design Conditions
BS 8002(36) *
CIRIA 104(66)
Effective
stress
Total
stress
1.2
1.5
1.2
1.0
These are mobilisation factors appropriate for a serviceability limit state and are not strength
factors as such.
Larger than 1.5 for clays which require large strains to mobilise peak strength.
Eurocode 7 provides partial factors for c'; N'; and cu; which are presented in
Table 10.2.
Table 10.2 Eurocode 7 partial factors for soil strength parameters
Case
Ground Properties
tan N'
c'
cu
1.1
1.3
1.2
B1
1.0
1.0
1.0
1.25
1.6
1.4
Case A
Case B
Case C
10.2.2
In this approach the earth pressure distributions are calculated using the fully
mobilised (unfactored) design soil strengths and the geometry adjusted such that
restoring moments exceed overturning moments by a prescribed margin. This is an
empirical method developed by Burland, Potts and Walsh which has been shown to
behave successfully. It is a consistent method providing satisfactory results using
one lumped factor of safety, Fr, for the practical range of soils and wall geometries.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
73
15 September 2003
The resultant earth pressure forces acting on the wall are split into net activating and
net resisting components. The net activating forces are those forces that arise from
the retained height of soil, whilst the net resisting forces are those forces from the
soil below excavation level (see Figure 10.1). These net forces are expressed as a
function of the unknown depth of embedment, do and their calculation involves the
solution of a cubic equation.
Earth
Passive
Net water
Active
Active
Figure 10.1 Earth pressure representation for the Burland & Potts method
The active earth pressure diagram is modified by altering the pressure at any level
below excavation to be equal to the pressure at excavation level. The passive earth
pressure diagram is modified by deducting the difference between the gross and
modified active pressures from the gross passive pressure at any level. The
modified active and passive pressure diagrams are given by the unshaded areas.
The design embedment is that which satisfies the following relationship:
M mp
Fr
Mmp
Ma
Mw
Fr
= Ma + M w
=
=
=
=
Care must be exercised in the use of this method where a value of c' is used when
analysing total stress conditions (see paper by Burland, Potts and Walsh(74)).
Recommended factor of safety values for the Burland and Potts method are
presented in Table 10.3.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
74
15 September 2003
Reference
Design condition
Fp
1.0
1.5
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
75
1.5 - 2.0
15 September 2003
11
For integral bridges the analysis of earth pressure requires a strain dependent model
that can only be provided by a numerical analysis computer program.
This Section provides guidance on the determination of bending moments and shear
forces acting on the bridge abutment where the mobilisation of earth pressures and
soil-structure interaction are considered. The soil-structure interaction approach
produces a more realistic representation of the behaviour of a structure by taking
into account in situ soil stresses, temperature effects, forces and bending moments
from the bridge deck and the stiffness of the structure and the soil. As these
analyses cannot be performed by hand calculations, commercially available software
is required.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
76
15 September 2003
Ko
Ka
Kp
K o = 2.0
1
4.0
H/2
1
3
Earth pressure
coefficient K
H
K o = 0.5
3.0
3
2.0
Ka
Active side
Kp
Passive side
1.0
4.0
3.2
2.4
1.6
0.8
0
/H (%)
0.8
1.6
2.4
3.2
4.0
Actual measured
pressure
Pressure profile
from classical theory
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
77
15 September 2003
11.3.1
Slider
Spring
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
78
15 September 2003
11.3.2
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
79
15 September 2003
11.5.1
WALLAP
The deliverables from the program include a stability analysis, wall displacement
versus depth profile, bending moment, shear force, earth pressure distributions and
strut loads.
Users need to be aware that the Winkler spring approximations do not yield the
same wall displacements as those given by more sophisticated models. See Brooks
and Spencer 1992(79), who compared the results from WALLAP and FLAC. At that
time they concluded that lower stiffnesses must be used in the Winkler spring
models to obtain similar displacements to those given by finite element models.
11.5.2
FREW
FREW (Flexible REtaining Wall analysis) is part of the OASYS suite from Ove
Arup and Partners, London which is also developed specifically for retaining wall
design. The program uses a modified Winkler spring model approach, see Pappin
et al(80). The analysis is carried out by assembling a stiffness matrix for a line of
nodal points that represents the wall. The soil can be modelled in three different
ways:
C
FREW calculates earth pressure, shear forces, wall bending moments, prop forces
and wall displacements for each construction stage being considered and also allows
soil arching to be modelled. Full details are given by Pappin et al(81).
FREW has the facility to model the effect of moment continuity between wall and
deck by adding deck elements as a continuation of the wall above the soil (that is,
as though the deck is rotated by 90E to make it vertical). The extent of the deck
elements and the restraints at its end are chosen to represent the whole deck. For
a simple span, only half of the deck is modelled; at the centreline restraints are
provided to simulate the affect of the complete deck on a pair of wall supports.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
80
15 September 2003
Symmetric and anti-symmetric loading can be modelled. The SCI publication Steel
integral bridges: Design of a single-span bridge - Worked example(5) provides an
example of the use of FREW.
11.5.3
FLAC
As yet there are no programs that allow a three dimensional soil-structure model to
be produced easily. Until such a program is developed, it will be necessary to
model the soil and structure separately, and approximate the interaction between the
two by the provision of appropriate boundary conditions. The commonly used
retaining wall analysis programs such as WALLAP or FREW are capable of being
used as part of such an analysis.
One possible analysis method is illustrated in Figure 11.4. The integral bridge is
split into two models, deck grillage and abutment. Boundary conditions are
provided at the connection between the two models. The rotational stiffness of the
combined wall and soil is calculated using the WALLAP model, by applying
moments directly to the top of the wall and measuring the resulting rotation. It is
suggested that the full fixed end moment is applied as an upper bound, in order to
establish that the soil response is linear within the expected range of moments. The
rotational stiffness of the deck can be established in a similar manner. The values
of the respective spring forces are transferred between models.
Spring A
Spring A
Deck Model
(Grillage)
Spring B
Spring B
Elasto
plastic
soil
mass
Soil Model
(WALLAP, FREW)
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
81
15 September 2003
For a beam of constant stiffness, the general stiffness/flexibility equation for in-plane
rotation may be expressed as:
6 EI y
My1 =
L2
d z1 +
4 EI y
L
y1 +
6 EI y
L2
d z2 +
2 EI y
L
y2
(1)
y1
my2
my1
y
x
M y1
y1
2 EI
L
Loading on the deck must be split into symmetric and anti-symmetric components.
The stiffness for symmetric components is 2EI/L, as above; the stiffness for
anti-symmetric components is 6EI/L.
The actual deck expansion depends on the point of equilibrium between the soil
reaction and the internal forces in the deck. This is illustrated diagramatically in
Figure 11.6.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
82
15 September 2003
Force
Deck behaviour
Soil behaviour
Force
at
equilibrium
Point of
equilibrium
Displacement
at
equilibrium
Displacement
In this Figure, the force-displacement line for the deck represents the response of
the deck to an axial load after a thermal strain has occurred. The intercept on the
force axis represents the force if all strain is prevented; the intercept on the
displacement axis represents the displacement at the end of the deck if it is free to
expand.
The force-displacement line for the soil mass can be determined using FREW or
WALLAP. It is suggested that preloads to the strut are applied up to the fully
restrained force in the deck.
The line represents the displacement at the top of the wall when subject to a load
applied at the top of the wall (additional to any preload due to the action of the deck
as a prop).
The intersection of the soil line with the deck line will be the actual force and
displacements for the given temperature increase.
If the two abutments are similar, there will be equal (and opposite) displacements
at each. If they are dissimilar, a more complex interaction must be considered.
11.6.2
The load combinations that involve braking loads need careful consideration of an
appropriate model because of the asymmetry caused. A sway type effect is created
with both frame and pinned abutment bridges.
In a frame integral bridge, this may contribute to producing critical bending
moments at the integral connection. To check whether the affects may be
significant, Kleinlogel graphs (see the Steel Designers Manual(82)) can be used for
a quick initial analysis, by simplifying the structure to a portal frame with a fixed
base - initially 3 m into stiff clay (after McShane(21)). Neither FREW nor WALLAP
are particularly suited to the direct evaluation of the effects of this loading since both
the loading and the structure (including the soil mass) are non-symmetric.
This check is illustrated in Steel integral bridges: Design of a single-span bridge Worked example(5), which calculates the moments at the wall/deck junction acting as
a portal frame without restraint from the soil.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
83
15 September 2003
12
This Section provides guidance in analysing the bridge abutment for forces resulting
from the cyclic expansion and contraction of the bridge deck beam due to
temperature effects.
The movement at the ends of the bridge due to change of temperature in the deck
is a significant factor in the design of an integral bridge, so it is important to have
a good understanding of the temperature changes that the bridge will experience.
The length of a bridge depends on its effective bridge temperature which, in
essence, is the mean temperature throughout the structure, or the temperature at the
neutral axis of the deck beams. The effective bridge temperature depends on
ambient temperature, on the type of construction, on the duration of the weather
conditions and the wind circulation. The determination of effective bridge
temperature has been related to ambient temperatures by Emerson(62). The maximum
and minimum ambient temperature (over the Highways Agency required 120 year
design life) depends on the geographical location of the bridge in the UK.
HA document BD 37(51) defines the requirements for the determination of maximum
and minimum effective bridge temperature. These temperatures are used to
calculate the change in length (of an unrestrained bridge) at the extremes of ambient
temperature variation, and hence the maximum conceivable movements for design
purposes. However, the neutral, or at rest position of an integral bridge can be at
a temperature which is not the mean of these two extremes. A practical judgement
assumes that the bridge is set at a temperature within 10EC above or below the
mean; the maximum movement is therefore that due to half the total range plus
10EC. This appears to be the basis used in BA 42/96(9), where a thermal strain
range is specified for the design of integral bridges.
Data on temperature ranges, nominal thermal strains and characteristic thermal
strain range according to BA 42/96 are presented in Table 12.1.
For integral bridges, BA 42/96 advises that no factor need be applied to the thermal
strain range; at ULS, earth pressures calculated on the basis of these strains are,
however, subject to a factor (fL of 1.5 (Clause 2.8 of BA 42/96). It may be noted
that this contrasts with BD 37, which specifies, separately, that ULS design values
are obtained by applying a factor ((fL) of 1.3 on the nominal range of movement,
or by applying a factor of 1.3 on the forces due to restraint of the nominal range of
movement.
For design purposes, if the thermal strain range is taken according to BA 42/96, i.e.
0.0005 for a composite bridge, without any partial factor, this corresponds to
movements at each end of a 60 m total length bridge of some 15 mm, assuming that
both ends move equally.
In practice, the movements observed at the ends of monitored bridges appear to be
much smaller than the design movements. This discrepancy is due to actual
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
84
15 September 2003
temperature variation, on a short-term basis, being much less than the extreme
design ranges. For example, Emerson(62) notes that the daily range in effective
temperature of a concrete bridge is less than 4.5EC for 96% of the days in a year
compared with the nominal design range of 46EC. The observed movements on
supposedly unrestrained bridges may be less than expected due to frictional restraint
in the bearings. Bridges in which bearings become locked up nevertheless
continue to expand and contract, displacing their supports as they do so. In these
cases simply supported bridges may actually be acting integrally.
Table 12.1 Thermal strain design temperatures
Deck Structure
type
Ambient temperatures
Temperature
o
C
Thermal
strain
Thermal
strain range,
in BA 42
Maximum
o
C
Minimum
o
C
Groups 1 & 2:
Steel
47
!28
47.5
0.00057
0.0006
Group 3:
Composite
40
!19
39.5
0.00047
0.0005
Group 4:
Concrete
37
!14
35.5
0.00043
0.0004
Notes:
1) The ambient temperatures are respectively the maximum and minimum anywhere in the UK,
so the range is greater than might be expected for any given location.
2) Structural types are: Group 1 steel box girders; Group 2 steel girders; Group 3 composite;
Group 4 concrete.
The behaviour of soil under cyclic loading is largely influenced by the rate at which
the change in pore water pressure can dissipate through the soil mass. In low
permeability cohesive soils the rate of dissipation is very slow and the soil usually
responds under undrained conditions to cyclic loading. Under these conditions the
undrained shear strength cu is reduced with increasing shear strain. In contrast, if
partial or full drainage of the soil occurs then consolidation and strengthening of the
soil takes place, together with an increase in stiffness.
Research has been undertaken on the behaviour of soils under cyclic loading
particularly where compaction of soils by mechanical plant occurs. See Section
12.3. However, cyclic loading compaction under surface moving loads will be of
relatively high frequency and in most cases there is little opportunity for soil
drainage and dissipation of pore pressures to occur even in relatively permeable noncohesive sands and silts.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
85
15 September 2003
Limited work has been undertaken and few cases published of the effects of thermal
movement on lateral stresses in integral bridge abutments. Broms and Ingleson(83)
measured lateral earth pressures acting on backfilled abutments of rigid frame
bridges. They found that complex interaction between abutment wall and its backfill
occurred during a small number of cyclic movements. Further work by Broms and
Ingleson(84)(85) indicated a continuing increasing trend in lateral earth pressures with
increasing number of cycles of bridge movement.
Backfilled abutments are generally speaking designed on the basis that the backfill
pressure equates to at-rest conditions, Ko, whilst the earth pressures for embedded
walls would in practice be the in situ, at-rest condition and the design pressures
would generally be assessed on the basis of active lateral pressures Ka. Where the
in situ soil is overconsolidated either naturally or by compaction, higher lateral earth
pressures theoretically up to Kp could exist. More recent work by Ingold on the A3
Wisley bridge over the M25 in Surrey has produced further data on earth pressures
in fill behind endscreen walls (see Section 14.4).
For embedded steel sheet pile retaining walls, the initial earth pressure mobilised in
the soil behind the wall is likely to be the Ko value, which is likely to be greater
than 1.0 in natural overconsolidated clays and probably of the order of 0.5 in sands.
The Ko value dictates the earth pressure for cuttings in natural clay and for the
design of retaining walls for deep urban underpasses it is advisable to measure the
in situ Ko value during the site investigation using such tools as the dilatometer.
Recently research by Springman et al at Cambridge University to study the effect
of slow cyclic temperature induced expansion or contraction on the displacement of
flexible sheet pile retaining walls in sand has been published in the TRL Report
Cyclic loading of sand behind integral bridge abutments(86). Based on the physical
tests and numerical analyses conducted in this research, actual lateral earth pressures
on the wall have been produced which provide a better insight into the actual
behaviour of the soil and the wall. These lateral earth pressures are presented in
Figures 12.1 and 12.2.
SCI studies(1) show that wall movements due to deck expansion for the range of
bridges with deck lengths less than 60 m, are insufficient to generate full passive
pressure and therefore this experimental research is not relevant. The worst case
condition used in the Springman et al research is to extreme to apply in most
integral bridge design and the designer is best advised to use methods explained in
Sections 11.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
86
15 September 2003
Depth (m)
Wall
static
cycle 1
2
K=2
4
e
activ
dr
hy
12
200
active
10
os
ta
tic
tic
ta
os
dr
st
h y at re
at re
st
100
K=2
-100
-200
Depth (m)
Wall
static
cycle 1
2
4
12
200
active
100
tic
ta
os
dr
hy at rest
10
hy
dr
os
ta
tic
at re
st
active
K=2
-100
-200
Figure 12.2 Lateral earth pressures - wall displacement away from fill
The results have shown that at serviceability limit state, the effect of cyclic passive
wall rotation on lateral pressures, bending moments and deck loads did not appear
to be significant at these amplitudes of wall rotation. However, at ultimate limit
state under passive wall rotations, abutment bending moment and axial deck load
values are noticeably increased by the stiffening of the soil response due to the
cyclic nature of wall movements. The lateral earth pressures on the abutment wall
are also increased due to ongoing densification of the retained soil.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
87
15 September 2003
Vertical
20E forwards
20Ebackwards
30E
35E
12
40E
20
45E
15
37
It can be seen that Kp increases very rapidly at high angles of friction and it is
therefore essential to have reliable measured values of N for the select granular fill
materials that are likely to be used immediately behind abutment walls for integral
bridges.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
88
15 September 2003
... the relationship between K*, the retained height (H) and thermal
displacement of the top of the abutment (d) [is]:
3.5.3
3.5.4
3.6 Live load surcharge on backfill should be ignored when calculating the passive
earth pressure mobilised by thermal expansion of the deck . Earth pressures
under live load surcharge are to be checked at the at-rest earth pressure
conditions with K0 = (1 - sin N').
2.9 Earth pressure coefficients on abutments should be calculated ... using material
factors (m on earth pressure coefficients of:
disadvantageous forces from backfill (m=1.0
The relationship given in 3.5.3 is illustrated graphically in Figure 12.3 for a typical
abutment height of 7 m and an endscreen wall of height 2 m. The lower limit
specified in 3.5.4 is also indicated in the Figure.
0.50
Endscreen wall
H = 2.0 m
0.45
0.40
Lower limit
0.35
K* / K p
advantageous forces from backfill when resisting secondary load effects (e.g.
braking) (m = 0.5
0.30
0.25
Typical abutment
H = 7.0 m
0.20
0.15
0.10
0.05
0.00
10
12
14
16
Displacement, mm
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
89
15 September 2003
Force (MN)
6.0
5.0
4.0
3.0
2.0
1.0
0.0
Response of wall
0
10
15
Displacement (mm)
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
90
15 September 2003
Kp
Ko
30
0.5
35
3.7
0.43
40
4.6
0.36
45
5.8
0.29
In BS 8002 N'design is determined from the definition given in Clause 3.2.5 where the
mobilisation factor M is equal to 1.2. Values of N'design are given in Table 12.4.
N'design
Kp
Ko
30
25.7
2.5
0.57
35
30.3
3.0
0.50
40
35.0
3.7
0.43
45
41.8
5.5
0.33
Note:
It can be seen from Table 12.4 that N'design and Kp are now lower than N'peak and its
corresponding value of Kp. This means that the design presumes a looser soil.
However, for integral bridges, the cyclic movement due to expansion and
contraction of the deck results in the compaction of the soil behind the retaining wall
abutment. BS 8002 does not cater for the compaction condition occurring. BA
42/96 attempts to allow for the compaction of the soil by using a mobilisation factor
of 1/1.2 rather than 1.2. It quotes that a factor of less than 1.0 can be allowed as
it reflects the approach given in Eurocode 7(48). The mobilisation factor of 1/1.2
increases both N'design and Kp which correspondingly reflect a compaction of the soil
behind the retaining wall abutment. The resulting values of N'design and Kp are given
in Table 12.5.
Table 12.5 Values of Kp and Ko given by BS 8002, using M = 1/1.2
(BA 42, Clause 3.3)
N'
N'design
Kp
Ko
30
34.7
3.64
0.43
35
40
4.6
0.36
40
45.2
5.9
0.29
45
50.2
7.6
0.23
Note:
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
91
15 September 2003
However, the values of Kp in Table 12.5 are not the same as those in the Table in
Clause 3.4 of BA 42/96. If the values of that Table are back-analysed, it is found
that the mobilisation factor is not 1/1.2 but is in the range 1/1.81 to 1/1.56
depending on the value of N' (see Table 12.6). Designers should seek clarification
from the Highways Agency before proceeding with their design.
Table 12.6 Mobilisation factor M (from definition in BS 8002 from Table
in BA 42/96 Clause 3.4
Note:
N'
Kp
N'
Mobilisation
factor M
30
42.0
34.7
35
45.6
40
40
53.1
45.2
45
15
61.0
50.2
Current practice for flexible walls is to use highly developed numerical analysis
computer programs such as FREW and WALLAP. These programs model the earth
pressure versus wall movement relationships and are accepted because they have
been calibrated by comparison to measured wall behaviour. It is also accepted that
current knowledge does not permit really accurate predictions of wall movement,
and differences of the order of 100% are common between predicted and measured
values or between the results of different programs!
Hence any method that is presented for the analysis of earth pressure should be
treated with caution and the advice of experienced practitioners sought in the process
of design judgement. Care is obviously necessary in design to properly understand
how a particular integral bridge will behave and routine design methods should only
be included where there are adequate checks.
In this light, the limits presented by BA 42/96 should not be taken as definitive,
more as an approach that attempts to be guarded and conservative if possible; the
objective of the Highways Agency is to prevent under-design.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
92
15 September 2003
13
This section presents methods for predicting the axial capacity of sheet piles and
bearing piles for bridge abutments or piers which resist vertical loads from the
bridge superstructure. The subject is thoroughly covered in the SCI publication
Steel bearing piles guide(12).
Axial loads acting at the top of the piled steel sub-structure are loads directly from
the bridge deck superstructure. These loads comprise dead loads, traffic loads and
environmental loads, and are transferred from the bridge deck to the abutment either
via bridge bearings or directly from the deck beam via the pile cap to the steel pile.
Finally the axial load in the pile is transferred to and resisted by the surrounding
soil.
Skin friction
resistance
End bearing
resistance
Figure 13.1 Wall friction and end bearing resistance against vertical loads
A simplistic relationship is given in BS 8004(43) for the ultimate capacity, R of the
pile. This relationship assumes that the ultimate capacity is equal to the sum of the
wall friction capacity Rsk and base capacity Rbk, i.e.:
Rc = Rs + RD = Asas + Abab
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
93
15 September 2003
where:
unit wall friction value 1)
surface area of the pile in contact with the soil 2)
unit end bearing value
steel cross section area at the tip of the pile or plug cross sectional area.
qs
As
qb
Ab
=
=
=
=
1)
The average value of qs over the length of the pile is taken for a soil profile
with more than one soil type.
See Section 13.7 for determination of surface area.
2)
The relative magnitudes of the ultimate wall friction and ultimate end bearing
resistances depend on the geometry of the pile and the soil profile. Where a pile is
embedded in a relatively soft layer of soil, but bears on a firmer stratum, this type
of pile is referred to as an end bearing pile. It derives most of its capacity from
the end bearing capacity, Rb. On the other hand, where no firmer stratum is
available to found the pile on, the pile is known as a friction pile. In cohesive
soils, the wall friction capacity Rs is generally paramount, whilst in non-cohesive
soils, the overall axial capacity is more evenly divided between wall friction and end
bearing capacity.
The equation presented above only considers the ultimate state condition, where the
pile has been allowed to deform sufficiently to allow both the ultimate wall friction
and the ultimate end bearing capacities to be developed. Commonly, load transfer
curves are produced which are plots of load resistance versus axial deformation of
the pile head for displacements ranging from zero to the ultimate limit or to an
achievable maximum value. These plots include mobilised soil-pile shaft friction
(shear) transfer versus local pile deflection and mobilised end bearing resistance
versus axial tip deflection. The Steel bearing piles guide(12) provides more detailed
information.
Numerous computer programs are available commercially to model the vertical
capacity of piles. One is PILE, which is part of the OASYS suite of geotechnical
programs(87).
where
Rsk
s
Rsk
Rbk
(s
(b
.
=
=
=
=
=
Rsk
b
ultimate wall resistance
ultimate bearing resistance
factor for wall friction resistance
factor for base resistance
factor to take into account uncertainty of soil parameters
determined on site or in the laboratory.
(s, (b and . are partial factors for the resistance side of the limit state equation.
These factors are not provided by BS 8002(36) or BS 8004(43) but are given in
DD-ENV 1997-1 Eurocode 7(48). In Eurocode 7 for driven piles:
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
94
15 September 2003
(s
(b
.
=
=
=
1.3
1.3
1.5
Provided that
PDesign
Rcd
where
PDesign is the design magnitude of the axial load including all appropriate
partial factors from BD 37(51), then the design vertical capacity of the sheet
pile - soil interface is adequate.
13.3.1
Wall friction
Most piles in clay develop a high proportion of their overall capacity in wall
friction, hence more effort has been devoted to developing reliable data for
estimating values of wall friction in clays than in sands. Previous work in this field
for both clays (and sands) is presented in API RP2A-LRFD(22) and various
background references are given. The unit wall friction qs for clays can be
estimated in terms of the undrained shear strength of the soil and is given in
BS 8004 and API RP2A-LRFD(22) by the relationship
qs = " cu
where
"
cu
=
=
a dimensionless factor
undrained tri-axial shear strength of the soil.
BS 8004 does not offer any specific advice on the value of ". However, SCI
studies(12) show that " can be conservatively taken to be 0.25 for the ultimate skin
friction value. (From pile load tests it has been found that the value of " increases
with time. In the short term (typically when site pile tests are performed) " is
approximately equal to 0.25 but with time (months/year later) " can reach a value
of 0.5.
13.3.2
End bearing
The long term drained end bearing capacity of a pile in clay is significantly greater
than its undrained capacity. However, the settlements required to mobilise the
drained capacity are far too large to be acceptable. Also, the immediate load
carrying capacity of a steel pile must be sufficient to support all loads during
construction. For these reasons it is common to calculate the base capacity of piles
in clay in terms of the undrained shear strength, cu.
The magnitude of the end bearing capacity qb, generated in cohesive soils is given
in BS 8004 as:
qb = 9cu
where
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
95
15 September 2003
Wall friction
where
2N
N
13.4.2
End bearing
Methods of estimating values of the end bearing resistance can be based either on
fundamental soil properties or soil properties determined directly from in situ
measurements. For cohesionless soils the most reliable method of predicting end
bearing resistance is to use the static cone penetrometer (Dutch cone) in the site
investigation. The end bearing resistance is calculated from the relationship:
qb
c
q
where:
c
q
is the average cone resistance within the zone influenced by stresses imposed
under the pile wall tip.
400N
For sands, the end bearing values are an order of magnitude greater than cohesive
soils and range in value up to 40 MPa. Although this value may seem high in
relation to the 10 MPa quoted in BS 8004, it is nevertheless realistic, and even
higher values (up to 70 MPa) have been measured in sands offshore.
96
15 September 2003
End bearing capacities for intact rocks are of a magnitude greater than end bearing
capacities for even cohesionless soils and end bearing resistance has been measured
in the range 100 - 400 MPa. In the case of weathered or highly jointed rock, end
bearing values reduce significantly and can be of the order 10 to 100 MPa.
Active soil
zone moving
downwards
Passive soil
zone moving upwards
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
97
15 September 2003
Excavation level
Assumed length
of wall providing
wall friction
resistance
Pile tip
13.7.1
The surface area of sheet piles and High Modulus Piles can be obtained from British
Steels Piling Handbook(24) in the section that lists the coating areas for piles. The
surface area per metre length of sheet pile can be taken to be 80% of the coated
area. For one face of the pile, use 40% of the coated area. This area is multiplied
by the length over which wall friction is mobilised.
Where it is found that the depth of embedment based on stability is insufficient to
provide the required vertical resistance capacity, it can be assumed that any extra
length of pile will have friction acting on both faces of the pile.
Similar calculations of surface area can be performed for High Modulus Piles and
box piles.
13.7.2
The surface area of tubular and box piles depends on whether or not a soil plug is
formed at the tip. If no plug is formed at the tip of the pile, the surface area is
given by the summation of outside and inside surface areas. If a plug is formed, the
surface area is based on the outside surface only.
13.7.3
H section piles
As for closed sections, the surface area of a H pile section depends on whether or
not a soil plug is formed at the tip. If no plug is formed at the tip of the pile, the
surface area is given by the total surface area of the H section. If a plug is formed,
the H Pile is assumed to be a closed box section of a size based on the external
dimensions of the H pile.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
98
15 September 2003
The area at the tip of the sheet pile acting in end bearing assumes that no soil
plugging is present. In this case, the area is given by the cross sectional area of the
steel.
For a High Modulus Pile, the composite steel cross sectional area is used, unless the
sheet pile is curtailed above the tip of the High Modulus Pile. In this case, the area
used is the cross sectional area of the universal beam only.
13.8.2
For tubular and box piles, the area to be used in the valuation of end bearing is the
full cross sectional area of the pile base comprising the pile wall and any soil plug.
The calculated ultimate pile end bearing across the whole cross section is compared
with the internal soil plug plus the pile wall tip end bearing and the lesser is taken.
13.8.3
H section piles
For H section piles, the area to be used in end bearing is the full cross sectional area
of the pile base comprising the pile wall and any soil plug. The calculated ultimate
pile end bearing across the whole cross section is compared with the internal soil
plug plus the pile wall tip end bearing and the lesser is taken.
13.9.1
The load resistance versus pile head displacement curves are very useful in
presenting the situation that exists at the ultimate condition and the working
condition. Figure 13.4 shows the mobilisation of resistance with deformation for
a cohesive soil, including wall friction, end bearing and their combination. For
simplicity the resistance profiles are drawn as straight lines rather than curves. It
is seen for this situation that at the ultimate state where the resistance is at a
maximum the pile head deflection is approximately 20 mm. However, to establish
the point of the curve which represents the working condition it is necessary to
review the partial factors that are used in the procedure to design for the ultimate
condition.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
99
15 September 2003
For a limit state design where partial factors for loads and resistance are used the
equation relating load and resistance at the ultimate limit state is given by:
P fl f3 =
where
P
(fL
(f3
Rcd
(
.
Rcd
=
=
=
=
=
=
Axial resistance
R cd
bk
(Ultimate)
R sk (Ultimate)
0.3 R cd
R bk (Ultimate)
End Bearing
10
20
30
40
50
Axial Displacement (mm)
1.4 11
. P =
Rcd
1.5 1.3
or
P =
Rcd
0.3 Rcd
3.0
At the serviceability limit state (working condition) the partial factors for load and
resistance are all equal to 1.0, therefore the working condition can be defined
accurately by the intercept of the curve at a resistance value of 0.3Rcd. It is seen
from Figure 13.4 that for an axial load magnitude of 0.3Rcd the pile head
displacement at the working condition, i.e. approximately 4 mm, is significantly
smaller than the pile head displacement at the ultimate condition, i.e. approximately
20 mm. In addition, it is seen that in the case of a cohesive soil the resistance is
predominantly provided by wall friction.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
100
15 September 2003
14
Steel piles can be subjected to two types of lateral loads - those acting at the pile cap
level and those acting on the pile shafts through the soil mass. Both of these types
of loading need to be taken into account in the design of piles.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
101
15 September 2003
The analytical methods discussed generally include lateral, vertical and moment
loading as separate cases. The limitations imposed by the available methods of
analysis are highlighted and guidance is given on the practical problems of assigning
realistic values to the necessary soil parameters, particularly concentrating on the
value for soil reaction stiffness. Recommendations are made concerning methods
of analysis which may be satisfactorily adopted in most circumstances. Sections
14.2.1 to 14.2.3, and 14.3 provide a summary of the analysis procedure.
14.2.1
The ultimate resistance of a laterally loaded pile can be estimated from approximate
solutions proposed by Broms(96)(97). These solutions are based on limiting
equilibrium principles to assess stability only. They are known to be conservative
but are recommended because the ultimate lateral resistance of a pile is not usually
the governing criteria. Solutions are available for piles in clays and sands. The
relevant parameters required for the analysis are related to soil strength (see
Section 9.1).
14.2.2
In most cases the P-Y curve form of the Winkler soil model is used because it is the
one most extensively validated. P-Y curves originate from instrumented lateral load
tests carried out on 762 mm OD tubular piles in the USA in the 1960s for offshore
design. The models of load resistance were derived from soil resistance
distributions required to match the bending stresses measured in the pile shaft strain
gauge instrumentation, i.e. curve-fitting to match bending moment diagrams. The
P-Y curve method is the only one in which it is possible to allow for significant
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
102
15 September 2003
cyclic loading of piles. This is useful for the structural design of the pile section but
does not give accurate displacements because the single piles had no head restraint.
It is explained in detail in the U.S. Offshore Design Code API RP2A(22) and in
computer programs.
Specific P-Y curves can be obtained for soft clays, stiff clays and sands from
References(98)(99)(100). Computer software(101) is also available.
A commercially available program to model a single pile is ALP. It is part of the
OASYS geotechnical suite of programs(68).
14.2.3
The capabilities of the various methods of analysis available to the engineer are
summarised in Table 14.1, as given in CIRIA Report 103(67).
Table 14.1 Summary of the output of method of analysis
Model
Limitations
Application
Output
Depth, slope,
moment and shear
of the pile at any
depth.
Winkler Medium
or
P-Y analysis
Elastic
continuum
A reasonable model for Single piles or pile groups under Output depends on
single piles or pile
working loads.
the particular
groups at working load.
program adopted,
Yield of the soil cannot
but typically
be included exactly.
includes deflection,
The limitations depend
slope, moment,
on the mathematics of
shear and axial load
the particular computer
distribution for each
solution chosen.
pile in the group,
Available programs are
and the overall
limited to constant or
stiffness and/or
linearly increasing soil
flexibility matrix of
modulus with depth.
the pile group.
The load resistance of a group of piles could be less than the sum of the
resistances of all the piles in the group acting independently.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
103
15 September 2003
The pile head deflection of the group, or of its pile cap, may be different to
that of a single pile.
a)
Load resistance
There are some simple spacing rules that have been derived from experiment and
experience and piles will only interact to cause a group effect if they are closer
than a predefined spacing to each other. This is about 3D (where D = pile
diameter) for piles in clay or within about 4D if the piles are in sands. When they
are closer than those limits, the pile group behaves as a single block with shaft
friction around its external periphery and a base resistance over the whole area of
the block because the individual soil resistance envelopes overlap.
Wherever possible the layout, spacing and pile cross section size should comply with
the above criteria in order to ensure that piles act independently. If this is achieved
then no pile group effect will occur and the vertical load resistance of the group is
the sum of the individual piles, and the vertical deflection at the pile head is no more
than that of an individual pile.
Where such adjustment of the spacing and arrangement of the piles in a group still
violates these rules, then a pile group analysis will be required to determine the
interaction effects.
As a first approximation, the piles in the group should be arranged to resist the
applied loading from a structural point of view (i.e. the centre of action of the pile
group should lie near the resultant thrusts of the various load cases). Having
outlined a trial pile group, detailed analysis is carried out to refine the design.
A well detailed procedure for analysis of pile groups is given in the CIRIA Report
Design of laterally loaded piles(67) and is reproduced in Figure 14.2. The report
suggests that three levels of appraisal are adopted:
1.
2.
3.
Load factors applicable to the Limit State design of a piled foundation subjected to
substantial lateral load are not well established. Selection of the appropriate factors
depends on the type of loading, the reliability of the ground investigation data, and
the response of the completed structure to the deformation of the foundation.
Inherent uncertainties in assessing the loads and stresses in a laterally-loaded pile
group require that reasonably conservative overall safety factors should be adopted,
combined with limits on permissible displacement. The following tentative
guidelines are suggested for the design of individual piles:
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
104
15 September 2003
Assess
- reliability of data
- sensitivity of structure
Static Analysis
Refine
- pile group
- size of piles
Analyse Data
Select design
parameters
Analysis
Single piles
1. Elastic continuum methods
2. Subgrade reaction methods (p-y analysis)
Pile
1.
2.
3.
Minor Structures
Appraise performance
groups
Poulos analysis
Randolph analysis
Elastic continuum computer models
(e.g. PGROUP, LAWPILE)
Output
Piles
- axial load
- shear forces
- bending moments
Group
- vertical deformation
- horizontal deformation
- rotation
- flexibility matrix of group
Modify design
as neccessary
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
105
15 September 2003
1.
An overall Factor of Safety for lateral load of not less than 3 should be used.
2.
Limits on lateral pile displacement at the ground surface of not greater than 2%
of the pile diameter for sands and stiff clays, and not greater than 5% of the
pile diameter for soft clays, subject to the tolerances imposed by the structure
itself.
These criteria are conservative, and they should restrict the development of plastic
failure zones to a shallow depth.
(b) Pile head deflection
Three conceptual models are currently in use for the design of pile groups:
1.
Structural frame model - In the static and stiffness methods, the piles are
implicitly assumed to be end bearing on a competent layer, and the contribution
of overlying soft material to load capacity is entirely discounted. (Computation
of the forces in the frame is carried out by conventional structural analysis).
Although this is not a realistic model of actual site conditions, the method has
given satisfactory results, and it is still extensively used. The method is now
principally useful for a preliminary appraisal of the layout of a pile group and
for the design of lightly-loaded groups. For an economic design of pile groups
subjected to large lateral loads or moments, other forms of analysis are
preferred.
2.
3.
Elastic continuum model - An elastic continuum model is useful for the analysis
of both single piles and pile groups when the soil can reasonably be assumed
to be linearly elastic. In practice, provided an appropriate secant modulus is
selected, the method gives satisfactory results for piles at working load in most
soil types. Complex elastic-plastic soil models for pile group analysis are not
generally available and are not necessary if the limit of 0.02D is used for lateral
displacement at ground level.
Continuum analyses use Mindlins(105) solution (e.g. by Poulos(106)(107)(108)(109)) and
by Bannerjee and Driscoll(110)) to incorporate pile/soil/pile interaction. The
method usually involves the use of a computer, and the designer should be fully
aware of the limitations of the particular program used. While the method is
currently best suited to the final design of the foundation, the publication of
parametric studies(111) makes the method more generally applicable. Several
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
106
15 September 2003
Analysis of the foundation at unfactored working loads enables the designer to assess
the significance of the computer predicted deflections and to include the stiffness
matrix of the foundation in the overall design of the structure.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
107
15 September 2003
Endscreen
wall
3.5m
P.T.F.E. sliding
bearing
Capping
beam
Column
4.5m
Base
1.0m
6.0m
A comparison of the actual pressure profiles throughout the depth of the abutment
with the original design assumptions is shown in Figure 14.4. The main feature of
the pressure profiles obtained was the sharp reversal in lateral pressure near the top
of the columns. This suggests an instantaneous centre of rotation of the abutment
at a point some 2 to 3 m above the slab base. The resistance of the compacted fill
to deflection imposed by the deck superstructure and to lateral pressures from behind
the capping beam was very effective. Consequently, the pressures generated by
compaction of the fill against the capping beam were far greater than the active
values assumed in the design (approaching a rectangular lateral pressure profile
rather than the traditionally assumed triangular profile). It is recommended that
these compaction pressures would be more realistically estimated using the methods
proposed by Ingold(115).
Height (m)
End screen
wall
Transverse
capping
beam
6
Resultant
4
Design
Pile
60
40
20
0
20
Lateral pressure (kPa)
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
108
15 September 2003
Since the mass of the bridge support is an important factor in the energy sharing
equation, it can be expected that reinforced concrete walls and barrettes are a
superior solution to reinforced concrete columns, but are, of course, much more
expensive to construct.
Steel piles can offer a superior crash resistant support element that will merely dent
under impact. The extent of denting can now be predicted with reliable accuracy
by energy sharing models that are used in the defence industry to study the effect
of projectiles on composite walls, but these have not been used to date in the bridge
design industry. Relevant research has also been carried out in the offshore industry
in the study of boat impact on offshore oil platform legs (see papers by Amdahl and
Frieze(116)).
Denting of steel tubular column-piles under collision loads can be prevented by
filling the tubular column with mass concrete, and with adequate restraint at the head
of the column will provide an impact-resistant support element. A steel tubular
column-pile is also very easy to install by driving and is thus a practical solution for
bridge supports. They deserve serious consideration in competition to reinforced
concrete walls and barrettes where cost-effectiveness is of interest and the confidence
in performance under collision loads is required.
Alternatively, steel H-piles can be used as the core of concrete columns to assist
crash resistance where the mass of the column or wall is considered adequate to
absorb the impact without shattering.
Further consideration and some guidance on the use of steel column-piles is given
in Integral steel bridges: Design of a multi-span bridge - Worked example(3).
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
109
15 September 2003
15
The design of the deck slab both in spreading wheel loads to the main beams
and as part of the effective section of the main beams.
The design of the shear connection between the steel girders and the deck slab.
In an integral bridge all these design considerations still have to be made, but in
addition the effects of the restraints at the ends, in introducing both axial load and
moment, now need to be evaluated. In previous Sections, the effect of the beam
stiffness on retaining wall behaviour (when there is moment continuity) was
discussed, together with the interaction between wall and deck axial stiffness (when
a temperature rise occurs in the bridge deck).
This Section highlights the aspects of the deck design that are affected by integral
construction.
110
15 September 2003
supports where there is moment continuity (a frame integral bridge), the effect of
axial stresses will be significant. In these regions the bottom flange is usually
braced so that it can work near to its maximum (factored yield) stress. Additional
axial stress would require additional flange area and probably more bracing (i.e.
further into the spans). Clearly, the benefit of the deck acting as a prop to the wall
would have to be balanced against a slight increase in girder size in these regions.
The introduction of a significant level of axial stress into regions of a beam that are
designed to resist a combination of moment and shear raises a query about the
applicability of clauses in BS 5400: Part 3(42). Clause 9.9.3 covers combined
bending and shear; Clause 9.9.4 covers combined bending and axial load; neither
covers a combination of all three types of force. But, since the axial force might
in some cases be a significant proportion of the resistance force, it would appear to
be unsafe to ignore it in Clause 9.9.3.
It is suggested that accounting for the axial force in addition to moment and shear
could be carried out simply by adding the term Pmax/PD to the expressions in 9.9.3.1
(c) and (d). Also, whilst the deck as a whole has a favourable slenderness ratio, the
limitation on PD will be the buckling of the bottom flange. This can be recognised
by treating the flange as the chord of a truss, determining its effective length in
accordance with Clause 12.4.1, and then using the slenderness of the flange to
determine the ultimate compressive stress from Figure 37. (This neglects the small
amount of restraint provided by the continuous web as part of a U-frame with the
deck, but it is conservative to do so.)
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
111
15 September 2003
16
REFERENCES
1.
2.
3.
4.
5.
THORBURN, S.
The interaction between bridges and embankments and the use of run-on
slabs
Henderson colloquium on Integral Bridges: What are they?
University of Surrey, September 1993, and
The use of run-on slabs - a solution or part of the problem?
Seminar on Design of Integral Bridges
International Association for Bridge and Structural Engineering (British
Group) at The Institution of Structural Engineers, London 23 January
1996
6.
7.
WALLBANK, E.J.
The performance of concrete in bridges - A survey of 200 highway
bridges
The Stationery Office, 1989
8.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 1, Section 3
BA/BD 57/95, Design for durability
The Stationery Office, 1995
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
112
15 September 2003
Discuss me ...
9.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 1, Section 3
BA 42/96 Design of integral bridges
The Stationery Office, 1996
10.
11.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 1, Section 2
BA 28/92 and BD 36/92, Evaluation of maintenance costs in comparing
alternative designs for highway structures
The Stationery Office, 1992
12.
BIDDLE, A.R.
Steel bearing piles guide
The Steel Construction Institute (to be published)
13.
14.
15.
HAMBLY, E. C.
Bridge design for durability
Integral bridge abutment details in practice and theory
TRL, 1992
16.
WASSERMAN, E.P.
Jointless bridge decks
AISC Engineering Journal, Vol. 24, No. 3, 93-100, 3rd Quarter, 1987
17.
18.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
113
15 September 2003
Discuss me ...
19.
20.
LOW, A.
Integral Bridges - Designers Viewpoint
Seminar on Design of Integral Bridges
International Association for Bridge and Structural Engineering (British
Group) at The Institution of Structural Engineers, London 23 January
1996
21.
McSHANE, G.
Steel sheet piling used in the combined role of bearing piles and earth
retaining members
Proceedings of the 4th International Conference on Piling and Deep
Foundations, Stresa, Italy, 7-12 April 1991
TESPA (Technical European Sheet Piling Association), 1991
22.
23.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 1, Section 3
BD 60/94, Design of highway bridges for vehicle collision loads
The Stationery Office, 1995
24.
25.
26.
27.
28.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
114
15 September 2003
Discuss me ...
29.
30.
YANDZIO, E.
Design guide for steel sheet piled bridge abutments
The Steel Construction Institute, 1998
31.
32.
33.
CORNFIELD, G.M.
Steel bearing piles, 4th Edition
The Steel Construction Institute, 1989
34.
35.
HIGHWAYS AGENCY
BD 42/94 Design of embedded retaining walls and bridge abutments
(unpropped or propped at the top)
The Stationery Office, 1994
36.
37.
38.
39.
40.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
115
15 September 2003
Discuss me ...
41.
42.
43.
44.
45.
AKROYD, T.N.W.
Earth-retaining structures: Introduction to the Code of Practice BS 8002
The Structural Engineer, Vol. 74, No. 21, 1996
46.
47.
48.
49.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 2, Section 1
BD 32/88, Piled foundations
The Stationery Office, 1988
50.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 2, Section 1
BA 25/88, Piled foundations
The Stationery Office, 1988
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
116
15 September 2003
Discuss me ...
51.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 1, Section 3
BD 37/88, Loads for highway bridges
The Stationery Office, 1989
52.
53.
54.
55.
56.
PECK, R.B.
Advantages and limitations of the Observational Method in applied soil
mechanics
Ninth Rankine Lecture
Geotechnique, Vol. 19, pp. 171-187, 1969
57.
CARDER, D.R.
TRL Report 172: Ground movements caused by different embedded
retaining wall construction techniques
Transport Research Laboratory, 1995
58.
59.
60.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
117
15 September 2003
Discuss me ...
61.
62.
EMERSON, M.
TRL Report 696: Bridge temperatures estimated from shade temperature
Transport Research Laboratory, 1976
63.
EMERSON, M.
TRL Report 765: Temperature differences in bridges: basis of design
requirements
Transport Research Laboratory, 1977
64.
65.
ILES, D.C.
Design guide for continuous composite bridges: 1 compact sections,
1989, 1993
Design guide for continuous composite bridges: 2 non-compact sections,
1989, 1993
Design guide for simply supported composite bridges, 1991
The Steel Construction Institute
66.
67.
ELSDON, W.K.
Report 103: Design of laterally-loaded piles
Construction Industry Research and Information Association (CIRIA),
1984
68.
OASYS
Geotechnical suite of programs for analysis and design
ALP - Laterally loaded pile analysis
Ove Arup & Partners (London), 1991
69.
70.
HIGHWAYS AGENCY
Design manual for roads and bridges, Volume 4, Section 1
HA 34/87: Ground investigation procedure
The Stationery Office, 1987
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
118
15 September 2003
Discuss me ...
71.
72.
73.
MARTIN, W.S.
CIRIA Special Publication 136: Site guide to foundation construction
Construction Industry Research and Information Association (CIRIA),
1996
74.
75.
WINKLER, E.
Die Lehre von Elastizitat und Festigkeit
Prague, 1867
76.
TERZAGHI, K.
Evaluation of coefficients of subgrade reaction
Geotechnique, Vol. 5, pp. 297-326, 1955
77.
78.
BORIN, D.L.
WALLAP anchored and cantilevered retaining wall analysis program:
Users manual (Version 4)
Geosolve (London), 1988
79.
80.
81.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
119
15 September 2003
Discuss me ...
82.
83.
84.
85.
86.
87.
88.
MEIGH, A. C.
Cone penetration testing
Construction Industry Research and Information Association and
Butterworths, 1987
89.
BOWLES, J.E.
Foundation analysis and design, 4th Edition
McGraw-Hill, 1988
90.
WANG, C.K.
Stability of rigid frames with non-uniform members
pp. 275-294, JSD, ASCE, Vol. 93, ST 1, Feb. 1967
91.
92.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
120
15 September 2003
Discuss me ...
93.
94.
95.
TOMLINSON, M.J.
Pile design and construction practice, 4th Edition
E & FN Spon, 1994
96.
BROMS, B.B.
Lateral resistance of piles in cohesive soils
Proc. Am. Soc. Civ. Engrs. - J. Soil Mech. Found. Div., Mar.
1964, 90 (SM2), 27-63
97.
BROMS, B.B.
Lateral resistance of piles in cohesionless soils
Proc. Soc. Civ. Engrs. - J. Soil Mech. Found. Div., May 1964, 90
(SM3), 123-156
98.
MATLOCK, H.
Offshore Technology Conference, Houston, Texas, USA
OTC 1204, Correlations for design of laterally loaded piles in soft clay
May 1970
99.
100.
101.
REESE, L.C.
Laterally loaded piles: program documentation
Proc. Am. Soc. Civ. Eng. - J. Geotech. Engng. Div. 103(GT4),
287 to 305, April 1977
102.
103.
HORSNELL, M.R.
Pile design - application of offshore technology to the onshore
environment
Proc. Syp. Computer Applications to geotechnical problems in Highway
Engineering, pp. 268-292, 1980
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
121
15 September 2003
Discuss me ...
104.
POULOS, H.G.
Users guide to program DEFPIG - deformation analysis of pile groups
School of Civil Engineering, University of Sydney, 1980
105.
MINDLIN, R.D.
Force at a point in the interior of a semi-infinite solid
Physics 1936 7 195
106.
POULOS, H.G.
Behaviour of laterally-loaded piles: single piles
Proc. Am. Soc. Civ. Engrs. - J. Soil Mech. Found. Div., May 1971,
97(SM5), 711 to 731
107.
POULOS, H.G.
Behaviour of laterally-loaded piles: pile groups
Proc. Am. Soc. Civ. Engrs. - J. Soil Mech. Found. Div., May 1971,
97(SM5), 733 to 751
108.
POULOS, H.G.
Design of pile foundations
Research Report 271, University of Sydney, School of Engineering,
1975
109.
POULOS, H.G.
Lateral loaded deflection prediction for pile groups
Proc. Am. Soc. Civ. Engrs. - J. Geotech. Engng. Div., Jan. 1975,
101(GT1), 19 to 33
110.
111.
112.
113.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
122
15 September 2003
Discuss me ...
114.
115.
INGOLD, T.S.
Lateral earth pressures on rigid bridge abutments
Journal of Inst. Highway Engineers, Vol. 26, No. 12, London,
(December 1979), pp. 2-7
116.
P:\CMP\Cmp657\pubs\P163\P163-Final.wpd
123
15 September 2003
Discuss me ...
Typeset and page make-up by The Steel Construction Institute, Ascot, Berkshire, SL5 7QN
Printed and bound by Staples Printers Rochester Limited, Rochester, Kent, ME2 4LT
1600 - 6/97