Integral Bridges 4b386t

P163: Integral Steel Bridges: Design Guidance Discuss me ...

PUBLICATION NUMBER P163

Integral Steel Bridges: Created on 06 February 2007 This material is copyright - all rights reserved. Use of this document is subject to the and conditions of the Steelbiz Licence Agreement

Design Guidance

A R BIDDLE BSc, CEng, MICE D C ILES MSc, ACGI, DIC, CEng, MICE E YANDZIO BSc, MEng, CEng, MIMarE

Published by: The Steel Construction Institute Silwood Park Ascot Berkshire SL5 7QN Telephone: Fax:

01344 623345 01344 622944

P163: Integral Steel Bridges: Design Guidance

Created on 06 February 2007 This material is copyright - all rights reserved. Use of this document is subject to the and conditions of the Steelbiz Licence Agreement

Discuss me ...

© 1997 The Steel Construction Institute Apart from any fair dealing for the purposes of research or private study or criticism or review, as permitted under the Copyright Designs and Patents Act, 1988, this publication may not be reproduced, stored, or transmitted, in any form or by any means, without the prior permission in writing of the publishers, or in the case of reprographic reproduction only in accordance with the of the licences issued by the UK Copyright Licensing Agency, or in accordance with the of licences issued by the appropriate Reproduction Rights Organisation outside the UK. Enquiries concerning reproduction outside the stated here The Steel Construction Institute, at the address given on the title page.

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 ts 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 s 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

Buckinghamshire County Council Ove Arup & Partners Tony Gee & Partners WS Atkins

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.


iii

15 September 2003



Discuss me ...


iv

15 September 2003


CONTENTS Page No.


SUMMARY

viii

1

INTRODUCTION

1

2

WHAT IS AN INTEGRAL BRIDGE? 2.1 Definition and terminology 2.2 Frame abutment integral bridges 2.3 Pinned integral bridges 2.4 Bankseat integral bridges 2.5 tless deck bridges 2.6 Additional considerations when choosing integral construction

2 2 2 3 4 5 6

3

WHY 3.1 3.2 3.3 3.4 3.5 3.6

4

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

5

SUBSTRUCTURES FOR INTEGRAL BRIDGES 5.1 Frame abutment integral bridges 5.2 Bankseat integral bridges 5.3 Pinned integral bridges 5.4 tless deck configuration 5.5 Intermediate s 5.6 Bearings 5.7 Skew bridges

21 21 25 26 28 29 31 33

6

STEEL SECTIONS FOR INTEGRAL BRIDGE PIERS AND ABUTMENTS 6.1 Continuous wall steel pile sections 6.2 Box piles 6.3 Tubular piles 6.4 H-Piles 6.5 Installation tolerances

35 35 37 38 39 40

CHOOSE AN INTEGRAL BRIDGE? Experience with non-integral construction Highways Agency requirements Advantages of integral bridges Practical aspects of the site that influence choice Whole life costing The use of steel piles in integral bridges


v

9 9 10 10 11 12 12

15 September 2003



6.6 6.7 6.8

Environmental factors associated with driven piles Driveability Corrosion allowances

41 41 42

7

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

8

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

9

SITE INVESTIGATION AND SOIL DATA FOR DESIGN 9.1 Soil data required for design 9.2 Site investigation 9.3 Selection and evaluation of soil parameters 9.4 Soil parameters for design of integral bridges

67 67 67 68 68

10

ABUTMENT WALLS - EMBEDDED WALL STABILITY 10.1 Cantilever and propped walls 10.2 Methods of analysis for stability against overturning

70 70 71

11

ABUTMENT WALLS - SOIL-STRUCTURE INTERACTION 11.1 Soil-structure interaction approach 11.2 Mobilisation of earth pressure and soil-structure interaction 11.3 Soil-structure interaction analysis methods 11.4 Global analysis of integral bridges 11.5 Available soil-structure interaction analysis software 11.6 Boundary conditions at the deck to abutment connection

76 76 76 77 79 79 81

12

ABUTMENT WALLS - RESPONSE TO THERMAL DECK MOVEMENTS 12.1 Bridge temperatures 12.2 Soil behaviour under cyclic loading 12.3 Earth pressures due to wall displacement

84 84 85 86


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

STEEL PILES - AXIAL LOAD RESISTANCE 13.1 Ultimate axial capacity and load transfer 13.2 Vertical settlement and serviceability 13.3 Ultimate capacity in cohesive soils 13.4 Ultimate capacity in cohesionless soils 13.5 Ultimate capacity in rock 13.6 Mobilisation of wall friction on a retaining wall 13.7 Determination of friction surface area 13.8 Determination of end bearing area 13.9 Buckling aspects of fully and partially embedded piles

14

STEEL COLUMN-PILES - LATERAL LOAD RESISTANCE 14.1 Lateral loads from soil 14.2 Lateral forces at pile head 14.3 Analysis of pile groups 14.4 Behaviour of a spill-through column-pile abutment 14.5 Integral bridges and crash resistance

101 101 101 103 107 109

15

COMPOSITE DECK DESIGN 15.1 Axial Loading 15.2 Moments due to frame action

110 110 111

16

REFERENCES

112


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 ts 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 s, 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 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 s.


The use of steel piling in the bridge s offers a ‘compliant’ structural element that is well suited to integral bridge construction. The behaviour of the steel s 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 s 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 intégral: guide de dimensionnement Résumé La construction de ponts de type “intégral” est actuellement en plein essor car elle permet d’éviter les problèmes de durabilité liés aux appuis mobiles des ponts á poutres traditionnels. La publication explique le concept de “pont intégral” et illustre les différentes configurations structurales qui peuvent être utilisées. Un point très important qui conditionne le bon comportement de ce type d’ouvrage est celui des déplacements provoqués, dans les appuis et le sol qu’ils retiennent, par les mouvements du tablier du pont dus aux variations thermiques. Il est indispensable que l’ingénieur projeteur soit bien au courant de ce problème et puisse le prendre en compte de manère correcte. Le guide couvre les points principaux du dimensionnement des ponts de type intégral et expose la méthodologie à suivre tant pour le dimensionnement du pont, pour les murs s situés aux extrémités du pont et pour les piles de ponts.


viii

15 September 2003


L’utilisation de piles en acier comme appuis du pont convient particulièrement bien pour ce type de constructions. Le comportement de ces s en acier est expliqué tant sous les charges provenant du tablier du pont que sous la poussée des terres. Les exxigences des autorités responsables des routes sont discutées et comparées à d’autres codes et règles de dimensionnement relatifs au comportement des sols. L’interaction entre les rigidités du tablier, des appuis et du sol est analysée et des recommendations sont formulées pour atteindre un bon comportement d’ensemble. Le guide fait référence à deux publications consacrées au même sujet et intitulées “Ponts en acier de type intégral: dimensionnement d’un pont à simple portée exemple d’application” et “Ponts en acier de type intégral: dimensionnement d’un pont à portées multiples - exemple d’application” qui illustrent de nombreux aspects couverts dans cette publication. Rahmenbrücken aus Stahl: Anleitung zur Berechnung


Zusammenfassung Der Bau von Rahmenbrücken wird aktiv verfolgt als ein Mittel, Probleme der Dauerhaftigkeit zu vermeiden, die sich bei gewöhnlichen Balkenbrücken infolge von beweglichen Auflagern ergeben. Diese Veröffentlichung erklärt den Begriff “Rahmenbrücke” und zeigt die verschiedenen statischen Systeme. Ein entscheidender Gesichtspunkt des Verhaltens einer Rahmenbrükke ist die Verformung im Bereich der Auflager und des gestützten Bodens infolge zyklischer, thermischer Dehnungen des Brückenbalkens. Brückenplaner müssen 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 Rahmenbrücken und die anzuwendende Berechnungsmethodik werden vermittelt, sowohl für Brücken mit Widerlagerwänden als auch für Brücken mit Auflagerbänken oder Auflagem aus Pfählen. Stahlpfähle für die Brückenauflager sind ein günstiges bauliches Element, das gut zum Bau von Rahmenbrücken paßt. Ihr Verhalten unter der Belastung aus dem Brückenbalken und dem Erddruck wird erklärt. Die Anforderungen der Straßenbaubehörde werden besprochen und mit anderen Vorschriften und Berechnungsregeln hinsichtlich des Bodenverhaltens verglichen. Die Interaktion zwischen der Steifigkeit des Brückenbalkens, der Auflager und des Bodens sowie die Anforderungen für die Verbindung zwischen den beiden, werden untersucht. Auf die begleitenden Publikationen “Rahmenbrücken aus Stahl: Berechnung einer einfeldrigen Brücke - Berechnungsbeispiel” und “Rahmenbrücken aus Stahl: Berechnung einer mehrfeldrigen Brücke - Berechnungsbeispiel” wird Bezug genommen; sie illustrieren viele der Aspekte, die in dieser Veröffentlichung behandelt werden.


ix

15 September 2003


Ponte integrali in acciaio: guida progettuale Sommario Il notevole interesse recentemente manifestato per il sistema costruttivo dei ponti integrali risulta motivato dalla possibilità di evitare, con riferimento ai tradizionali sistemi di ponte a travata, i problemi di durabilità imputabili agli spostamenti dei giunti. Questa pubblicazione introduce il significato di “ponte integrale” e presenta le varie tipologie strutturali che possono essere utilizzate. Un aspetto peculiare del comportamento di ponti integrali è rappresentato dal fatto che gli appoggi da ponte, unitamente al suolo che li sostiene, non sono soggetti agli effetti provocati dalle escursioni termiche dell’impalcato del ponte. I progettisti di ponti dovranno di conseguenza essere in grado di trattare la risposta del terreno e degli appoggi della struttura in relazione a tali spostamenti, sviluppando quindi esperienza in questo nuovo settore.


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. L’uso di pile in acciaio per l’appoggio 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 dall’impalcato e delle azioni esercitate dal terreno. I requisiti di queste strutture imposti dagli enti preposti alla viabilità sono discussi e paragonati con altri criteri generali e con regole di dimensionamento legate al comportamento del terreno. L’interazione tra la rigidezza di impalcato, appoggi e terreno è 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 più campate - esempio applicativo”, le quali trattano molti degli aspetti affrontati in questa guida progettuale. Puentes de acero integrales: Guía de Proyecto Resumen Actualmente la construcción 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 publicación explica lo que se entiende por “puente integral” e indica las tipologías estructurales utilizadas. Un aspecto clave en el funcionamiento de un puente integral es el desplazamiento impuesto por las deformaciones térmicas cíclicas del tablero a los apoyos y al suelo retenido por aquellos. Los proyectistas de puentes deberán 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 así como sobre la metodología que debe seguirse tanto para puentes con estribos P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

x

15 September 2003


de contención 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 construcción integral. Su comportamiento bajo las cargas del tablero y de la presión del suelo se explica cuidadosamente. También se analizan los requisitos establecidos por la Highway Agency, comparados con otras Normas y Reglas de buena práctica relativas al comportamiento del suelo. Se estudia la interacción entre las rigideces del tablero, soportes y suelo, así como los requisitos de unión entre aquéllos. A lo largo del trabajo se hace referencia a la publicaciónes 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 cálculo ponen de manifiesto muchos de los temas contenidos en esta obra.


Ändskärmsbroar i stål: Dimensioneringsvägledning Sammanfattning Ändskärmsbroar har börjat användas i allt högre utsträckning för att undvika de underhållsproblem som är förknippade med rörelsefogarna i traditionella stålbalksbroarna. Denna publikation förklarar verkningssättet och olika konstruktionslösningar för ändskärmsbroar. En fråga som tas upp är vad som händer när bron utvidgar och drar ihop sig i längsled, p g a temperaturändringar. Brokonstruktören får här lära sig att hantera jordtryck och stödkonstruktioner samt allmänt bygga upp kunskapen om denna konstruktionstyp för stålbroar. Det ges vägledning i dimensioneringsförutsättningar och dimensioneringsgång för ändskärmsbroar med olika typer av upplag. Användandet av stålpålar som broupplag erbjuder en konstruktionslösning som är väl lämpat för ändskärmsbroar. Det redogörs för hur stålfundamentet påverkas av laster från brodäck och jordtryck. Brittiska Vägverkets krav behandlas och jämförs med andra standarder och dimensioneringsregler rörande jordtryck. Interaktionen mellan förstyvningarna av brodäcket, brostöden och marken är utforskad och kraven på samverkan mellan dem är utredda. Hänvisningar görs även till publikationerna “Steel integral bridges: Design of a single-span bridge - worked example” och “Steel integral bridges: Design of a multi-span bridge - worked example”, som illustrerar många av de aspekter som omfattas av denna publikation.


xi

15 September 2003



Discuss me ...


xii

15 September 2003


1 INTRODUCTION A modern beam-type bridge comprises two essential structural components - a deck to span the gap and the s. 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 ts between the ends of the deck and the approach structures, and to provide simple bearings on the ing structures. The object has been not to constrain the thermal expansion or contraction of deck beams nor to restrain their rotation at s.


Multiple span bridges can then be formed from a series of simply-ed beams, with similar separation ts between the ends of the separate decks - this was particularly popular in construction in reinforced concrete. However, structural continuity over intermediate s has always been easy to achieve with composite decks, and this has little effect on the structure, other than a redistribution of the vertical reactions, but it does afford some economy in the deck construction. The consequence of providing simple details and separation ts 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 t must be provided, so that the gaps are never large enough to cause a hazard to the road s. The structural form of a beam-type bridge with movement ts 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 ts, both over intermediate s and at end s, 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 ts, 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 s is explored. Steel piles offer a degree of flexibility at s 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.


1

15 September 2003


2 WHAT IS AN INTEGRAL BRIDGE? 2.1

Definition and terminology

Generally, an integral bridge is one where the bridge deck is made without any movement ts at the abutments or between spans. The use of the ‘integral bridge’, ‘integral abutment’ and ‘integral construction’ have not been consistent to date and the extent of the ‘integrity’ between the deck and ing structure varies. To avoid confusion a more rigorous system of definition is required. Bridges without movement ts can be conveniently divided into two basic classes, termed ‘integral’, and ‘tless 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 tless deck bridges to describe the stub wall at the end of the deck that retains the adjacent road construction.

2.1.1 Integral bridges An ‘integral bridge’ is one that has structural continuity between the deck and the structural elements that it. There is no relative translational movement at any interface between the deck and the ing structure. Three forms of integral bridge are described in this publication - frame abutment, pinned and bankseat, described in Sections 2.2, 2.3 and 2.4 respectively.

2.1.2 tless deck bridges A ‘tless deck’ bridge differs from an integral bridge in that movement bearings are provided between the deck and the substructure that s it, ensuring that the ing elements are not subject to displacement as a result of thermal expansion/ contraction or of deflection under load.

2.1.3 s Integral bridges can have wall abutments, column piers or bankseat s 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

Frame abutment integral bridges

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 ing elements carry the axial loads which are the end shear forces from the deck beams.


2

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 ing 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

Integral bridge - displacements due to expansion

Figure 2.2

Integral bridge - displacements due to vertical loading

No intermediate is shown in these Figures. A configuration with intermediate s would behave in a similar manner at the end s and the principles illustrated would not be affected. For further comments on the behaviour at intermediate s, see Section 2.6.4.

2.3

Pinned integral bridges

In a frame abutment integral bridge, displacements due to temperature (thermal strains) and load on the deck induce reverse curvature at the head of the abutment wall, as shown in Figures 2.1 and 2.2, and adequate moment capacity is required in the connection. This can involve a complex reinforcement detail in reinforced concrete endwalls and capping beams to ensure moment and force transfer. The introduction of a ‘pin’ at the connection between the deck beams and an abutment removes these large hogging moments at the bridge end. This can be efficient where there is judged to be little to be gained from an integral connection. Such a pin can be achieved with a relatively simple pinned bearing; the detail of how this may be achieved is discussed in Section 5.3.


3

15 September 2003


The effects of introducing a pin on the deflections of the bridge structure is illustrated diagrammatically in Figure 2.3.

Figure 2.3

2.4

Pinned integral bridge - retaining wall displacements due to deck thermal expansion

Bankseat integral bridges


A bankseat structure is a common detail for highway bridges. A bankseat can be made part of an integral bridge by fully connecting it to the deck to make them structurally continuous. Since a bankseat only ‘sits’ on the soil, the structure foundation will move relative to the ground as a result of thermal expansion and contraction, and can rotate under deck loading. Such a bridge can be formed by an endscreen wall (across the ends of the deck beams) that has a footing foundation, thus combining the functions of vertical to the deck and lateral to the abutting road construction. A diagrammatic illustration of the deflections due to temperature and load effects on a bankseat integral bridge is given in Figure 2.4 and Figure 2.5.

Figure 2.4

Bankseat integral bridge - displacements due to expansion

Figure 2.5

Bankseat integral bridge - displacements due to vertical loading


4

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 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

tless deck bridges

As explained in Section 2.1.2, a tless deck bridge eliminates movement ts at the road surface, but the s are not integral with the deck structure. However, like a bankseat integral bridge, an endscreen wall is formed across the ends of the beams, presenting a vertical face to the abutting road construction.


An arrangement of a tless deck bridge on a bankseat is shown in Figure 2.6. Vertical loads are carried directly in bearing onto the soil. There is a movement interface between the deck beams and the , arranged as a narrow horizontal gap under the endscreen wall (see Section 5.4 for further details).

Sliding bearing

Figure 2.6

tless-deck bridge - bankseat

An arrangement on a piled is shown in Figure 2.7. A suitable detail would need to be provided at the bottom of the end wall, which moves relative to the ground beneath, so that any drainage water is conducted away from the piles.

Sliding bearing

Figure 2.7

tless-deck bridge - piled

Like conventional non-integral bridges, access must be provided to permit inspection of the bearings during the life of the bridge and provision made in the construction to allow for jacking up the deck for bearing replacement. The principal feature is therefore the absence of deck ts.


5

15 September 2003


2.6

Additional considerations when choosing integral construction

2.6.1 The deck-end/road construction interface The degree of restraint provided by the soil and road construction against an endscreen face to a deck, to the retaining wall or piles, may reduce but can never eliminate expansion and contraction movements due to thermal strains in the bridge deck. Indeed, studies(1) show that a significant proportion of the ‘free’ thermal movements still take place in an integral bridge deck. It is therefore futile to try to completely restrain such thermal movements by attempting ‘rigid’ abutment design.


Where the bridge deck length is small, thermal movements can be accommodated at the bridge ends by an asphaltic plug t in the road surface immediately behind the end of the bridge. The t 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 t area, allowing salt-laden runoff water to percolate down the back of the structure. Opinions vary as to the effectiveness of asphaltic plug ts 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 ts. 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 t at the junction of the endscreen wall and the road construction.

2.6.2 Approach slabs The problems described in Section 2.6.1 can be moved away from the end of the main deck structure and s by the use of an approach slab and this has been a standard detail in the USA. However the practice has had mixed success on some bridges in the USA(4) and in Scotland(5). Failures appear to be mainly due to the inadequacy of the connection between the slab and the end wall. Clearly, such a slab must have structural ties to the deck end and be properly keyed to the structure so that they move together, and will need to be designed with sufficient strength in bending that it can span over any local settlement due to traffic vibration. Thermal deck movement can cause longitudinal compression in the approach slab or associated heave of the underlying fill due to lateral displacement of the endscreen wall or abutment wall, and this can compound the other problems identified and explained in the above references(3). The arrangement of an approach slab is shown diagrammatically in Figure 2.8.


6

15 September 2003


Asphaltic plug t

Road surface

Road base Approach slab

Select granular fill

Integral bridge deck

Embankment fill Steel bearing piles

Figure 2.8

Typical arrangement of an approach slab as used in the USA


The maintenance problem at the t 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 t 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.

2.6.3 Retained and unretained approaches The clear opening provided by a bridge or by the end span of a multi-span bridge may either be bounded by a vertical face, or the ground beneath the bridge may be sloped upward to the underside of the bridge. Both arrangements are shown in the Figures in this Section and either can be used, regardless of whether the bridge is approached by embankments or spans across a cutting. When the opening is to be bounded by a vertical face, the length of deck is minimised, but a retaining wall is needed to the road formation and the underlying fill or the soil face in a cutting. Conventional arrangements are to provide either a reinforced concrete wall or a steel sheet pile wall and to the deck beams on top of the wall capping beam. When a retained configuration is to be incorporated into an integral bridge, there is a choice as to whether the ing and retaining functions are separated or combined. If they are combined, the head of the retaining wall is subjected to deck thermal displacements and this complicates the retaining wall design.


7

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 tless 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.

2.6.4 Intermediate s


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 s, but this arrangement is already quite normal for composite construction. However, there remains the question of whether the behaviour of intermediate s has any effect on the overall ‘integral’ behaviour. If the intermediate s 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 s are pinned to the beams (but still with no rotational restraint) - perhaps in order to offer restraint to the against collision loads - then they will be displaced by deck thermal strains unless it is the central 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

Displacement of intermediate s pinned to deck beams

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 in the column design.


8

15 September 2003


3 WHY CHOOSE AN INTEGRAL BRIDGE? 3.1

Experience with non-integral construction


Over the past thirty years, engineers have become more aware of the pitfalls associated with the use of expansion ts and expansion bearings. C

ts are expensive to buy, install, maintain and repair; repair costs can be as high as replacement costs.

C

Successive carriageway repaving will ultimately require that ts be replaced or raised.

C

Even so-called ‘waterproof’ ts will leak over time, allowing runoff water, often salt-laden, to penetrate through the t and thus accelerate corrosion damage to girder ends, bearings and ing substructures.

C

Accumulated dirt, stones and rubbish may fill recesses, which can, for example, lead to failure of elastomeric bearing glands.

C

Hardware for ts can be damaged and loosened by snow ploughs and heavy traffic.

C

Bearings are expensive to buy and install and costly to replace.

C

In time certain types of steel bearing may tilt and/or seize up because of loss of lubrication or build-up of corrosion.

C

Elastomeric bearings can split and rupture as a result of unanticipated movements, or ratchet out of position.

C

Seized expansion ts and malfunctioning expansion bearings can also lead to damage of the main structural .

In 1985 in the USA, a survey carried out by the Federal Highway istration found(6) that 75% of the bridges built using expansion ts 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 ts 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 t leakage. Cracking was the other main problem, particularly that due to early thermal effects.


9

15 September 2003


3.2

Highways Agency requirements

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 ts for expansion or contraction”. That Standard is concerned mainly with the Agency’s principles of design; details of the Agency’s 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

Advantages of integral bridges


Clearly, the first advantage is the elimination of the cost of, and additional work associated with, the provision of movement ts 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 ts, even when the extra work associated with ensuring structural continuity were taken into (see Burke(10)). The benefits of reduced maintenance costs and reduced risk of damage arising from leaky ts 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 s 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.


10

15 September 2003


Tolerance requirements reduced The close tolerances required for setting expansion bearings and ts are eliminated. (Although other tolerance requirements may be introduced, depending on the connection detail). Greater end span ratio ranges It is normal practice for non-integral bridges to limit the ratio of the end-span length to that of the adjacent span to approximately 0.6, to avoid the occurrence of uplift conditions under extreme loading; if uplift can occur, expensive hold-down arrangements will usually be needed. The continuity at the ends of a fully integral bridge automatically provides uplift restraint; even a compact integral bankseat detail acts as an additional counterweight. Other advantages are listed by Burke(10), such as the elimination of the risk of leakage at the ends, which is of particular benefit with girders of weathering steel. Such girders in non-integral bridges commonly need to be painted at the ends for added protection.


3.4

Practical aspects of the site that influence choice

The main practical factors influencing the choice of an integral configuration are: C

The overall length of the bridge.

C

The geometrical arrangement of the end s (e.g. whether there are side slopes, cut or fill earthworks).

C

The type of soil(s) on which the bridge is founded.

C

The practicality of replacing bearings and access for maintenance.

C

The retained height of fill.

C

The construction method to be used.

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 elements of the integral bridge that can also minimise site occupation time. When electing for an intermediate to the bridge, the need to make use of the deck to provide restraint to the top of the 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 s. In addition, a more economic substructure may be possible by considering compliant steel piled s.


11

15 September 2003


3.5

Whole life costing

In BD 36/92(11), the Highways Agency requires that, in comparing alternative designs for bridges, the ‘whole life cost’ is considered, i.e. the capital cost plus the direct and indirect costs of maintenance throughout the life of the structure. Since this Standard predates both BD 57/95(8) and BA 42/96(9), there is no mention of integral construction in it, but it would seem sensible that the choice between integral and non-integral construction should be considered in the same way.


The savings in ‘whole life cost’ by choosing an integral bridge concept can be: C

Reduced construction cost.

C

Elimination of the cost of maintenance and replacement of movement ts.

C

Elimination of the cost of maintenance and replacement of bearings (where they are omitted, or a reduced cost where simpler bearings are used).

C

Reduced allowance for the maintenance cost for the deck slab (because the worst potential source for deck reinforcement corrosion is eliminated).

Note that it is not entirely necessary to eliminate bearings to reduce maintenance costs. For example, simple steel rocker or knuckle bearings provide an effective pinned connection, but do not need anything more than an occasional clean and repaint as there are no ‘moving parts’ that can seize up. Further information is given in Section 5.6.

3.6

The use of steel piles in integral bridges

Steel piling can be used for elements in substructures and foundations for integral bridges, such as: C

Bridge abutments.

C

Intermediate piers.

C

Wing walls.

C

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

Construction is significantly quicker when compared to in situ reinforced concrete foundations.

C

There is no requirement to excavate for foundations.

C

There is no disturbance of the existing ground during piling.

C

The steel components are shop quality not site quality.

C

Piles can easily be made aesthetically pleasing.


12

15 September 2003


C

Piles can be placed in advance of other works.

C

Piles have immediate load carrying capacity.

C

Sheet pile walls provide a curtain walling to contain the working site.

3.6.1 Structural advantages of piling in integral bridges For integral bridges, steel piling provides a sufficiently stiff but flexible structural element in the substructure for integral bridges. It offers a compliant foundation that will not crack in bending and that can reduce retaining wall bending moments, relative to a rigid reinforced concrete alternative. ‘Compliant’ is a term that neatly summarises the characteristics of steel substructures that are beneficial to integral bridge behaviour.

3.6.2 Types of bridge with steel piling Different types of frame abutment, pinned and bankseat integral bridge are illustrated in Figure 3.1.


Frame

Pinned

Carriageway

Carriageway

Retaining wall

Retaining wall

Tubular column-pile pier


Figure 3.1

Bankseat

Piled bankseat

Types of integral bridge with steel piles

Some types of tless bridge with steel substructures are illustrated in Figure 3.2.


13

15 September 2003


Cantilever/Propped

Bankseat

Sliding Flexible seal

H piles

Carriageway

Retaining wall



Figure 3.2

Types of tless bridge with steel piles

3.6.3 Appearance of piled substructure of integral bridges Steel piles for bridges can be very dominant features on the urban and rural landscape. Careful design can make a considerable improvement to their appearance without leading to a significant increase in cost. Apart from having to satisfy the functional requirements, steel piling can be made to blend in with its surrounding environment as far as possible and to be aesthetically pleasing. The aspects that are important are: C

Height of abutment/pier and inclination of its front face.

C

Anchorages in the face.

C

Wing wall angle of return affecting the elevation of the wall.

C

Gradient and surface treatment of the adjacent ground.

C

Surface textures and paint colour of the facing walls, and the expression and position of vertical and horizontal construction ts.

C

Concrete footing walls.

C

The coping/capping beam of the abutment.

The appearance of a pile can be improved by providing features in the finished face or by decorative facings or claddings.


14

15 September 2003


4 HISTORICAL BACKGROUND 4.1

Integral bridges in Europe and the USA

Bridges where ts 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 ts, 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 ts 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.


4.1.1 Examples of integral bridges in the USA In the USA, the move towards ‘integral’ and ‘tless’ construction has led to a much greater use of continuous construction for multiple spans (it is now favoured by over 85% of state transportation departments) and to the design of ‘integral abutments’ (fully integral and semi-integral bridges in our terminology). Experience on the performance of integral bridges constructed in the USA is reviewed in a paper by Burke(4). The terminology used to describe the types of integral bridge in the USA, however, differs from the definitions proposed in Section 2 of this publication. Although similar concepts have been used, the lack of rigour in definition of bridge type can lead to some confusion. Generally, the US design of integral abutments appears to have been somewhat empirical, based on what is judged to have worked satisfactorily before, rather than by rigorous analysis. The latter would demand more understanding of the behaviour of such structures so that it can be incorporated into a design procedure. Most of the integral abutment details described in journals and other published papers involve the use of piles (normally steel H piles) to carry the vertical reaction. A typical configuration is that used in Tennessee(13), as shown in Figure 4.1. Steel bearing piles are cast into the abutment and provide flexibility under lateral displacement.


15

15 September 2003


Reinforcement Prestressed concrete beams

Selected granular fill

Bearing

Drainage aggregate

H piles

Figure 4.1

Tennessee integral abutment detail


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

Freysinnet concrete hinge detail (ref. Ohio detail(15))

A region of potential weakness in a variety of integral abutment details is the keying of the abutment to the deck slab or to the approach slab. Failures of inadequate details have been reported in the USA(16). The evidence suggests that the design of the connection must include very careful design and detail of the tying reinforcement. To do this effectively requires a better fundamental understanding


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 ing 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

Examples of integral bridges in the United Kingdom

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 1960’s. 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- overbridge; that all used steel piled s. 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).


17

15 September 2003


C L

19m

CL Canal

600 Ø tubular bearing piles @ 1700 centres

Figure 4.3

Stockley Park canal bridge


A41 Stone Bridge, Aylesbury, Buckinghamshire The replacement to the A41 Stone Bridge just north of Aylesbury (Figure 4.4) has a single span of 18 m over the River Thame. The new bridge was designed as an integral bridge. The integral bridge arrangement permitted increased freeboard to the river when in flood, whilst maintaining the existing vertical alignment of the road over a low floodplain. This avoided the further expense of associated roadworks. To provide the required freeboard, it was necessary to minimise the deck beam construction depth to just 600 mm. This was achieved by mobilising abutment fixed-end-moments to reduce the bending moments at mid-span. The deck was propped at mid-span during its construction and then connected to moment-carrying in situ concrete capping beams on sheet piled abutments. Permanent High Modulus steel sheet piling was chosen to provide a compliant retaining wall foundation of adequate stiffness and to enable a practical construction procedure that permitted construction in the dry with minimal excavation in the soft alluvial soils. A paper by S. Griffiths of Buckinghamshire County Council describing design and construction aspects of the replacement bridge is to be published in The Structural Engineer in 1997. 21490 In-situ reinforced concrete deck slab

Precast deck beams Concrete pile cap

High modulus pile web formed from 914x305 UB welded to sheet pile facing wall 17900

Figure 4.4

A41 Stone Bridge, Aylesbury

A134 Chad Brook Bridge, Suffolk Another example of a replacement integral bridge is that constructed over Chad Brook in Suffolk (Figure 4.5). In this case the bridge span is approximately 11 m. The configuration of the bridge is similar to the A41 bridge in Aylesbury in which steel sheet piled abutments are provided and where the pile head is integral with the bridge deck end capping beams. However, on this site steel box beams were provided for deck beam . The sheet piling permitted construction in the dry


18

15 September 2003


and the span to be reduced to a minimum. Further information pertaining to this bridge can be found in a paper(21) by McShane. 10.90m square span

Road level

Reinforced concrete deck

Grass block paving to stream banks Extent of sheet pile wing wall

1

Extent of bank facing

2 Box piles forming portal

Figure 4.5

A134 Chad Brook Bridge, Suffolk


Bridgend-By- The Bridgend-By- 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

Offshore experience with tubular hollow sections

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 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

Instrumented pile load tests in both granular and cohesive soils.

C

Improved soil investigation methods and tools, e.g. Dutch cone T’s.


19

15 September 2003


C

Analysis of pile behaviour and calibration of design prediction methods.

C

International pooling of knowledge and best practice.


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).


20

15 September 2003


5 SUBSTRUCTURES FOR INTEGRAL BRIDGES Traditional non-integral bridges have generally used reinforced concrete for the s of both composite steel/concrete and reinforced/prestressed concrete deck construction. Where piles are needed, both steel and concrete have been used beneath abutment walls, bankseats and for intermediate s. For integral and tless deck bridges, both types of deck construction will need different configurations to respond to the need to accommodate displacements due to thermal strain and the bending of the deck. This Section presents a variety of configurations for s to each of the types of integral and tless deck construction, with particular emphasis on the use of steel in the substructure, since that offers a versatile and compliant form that is well suited to efficient integral construction. Outline details of s and connection zones are illustrated.


5.1

Frame abutment integral bridges

As explained in Section 2.2 this type of bridge is essentially a simple portal frame, and this means that the end s will be subject not only to bending due to pressure of retained soil and axial load as a result of ing 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 s, primarily for frame abutment and pinned integral bridges, but also for other types of , e.g. steel column-piles.

5.1.1 Steel pile retaining walls Steel sheet piling is commonly used for retaining wall construction in bridges, and is competent to carry the vertical and lateral loading. Where the wall is subject to significant bending, High Modulus Piles (UB sections welded to Frodingham sheet piling - see Section 6) can provide increased stiffness and strength. For both types of wall the vertical loads and moments are introduced at the top of the wall via a suitable connection detail. For frame abutment bridges, where significant pile head bending moments will be introduced by a pile capping beam, High Modulus Piles may be needed for all but the shortest of bridges.


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 t


Concrete deck

Deck beam

Reinforced concrete pile capping beam/wall

Construction t Note: Similar arrangement for reinforced concrete deck beam

High Modulus Pile Sheet pile

Figure 5.1

UB

A built-in connection for a High Modulus Pile retaining wall

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).)


22

15 September 2003


The connection is formed by using a construction t 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

Effective load distribution from individual deck beams to a group of piles.

C

Accommodation of a deck girder or beam spacing that is different to the spacing of the UB sections of the High Modulus Piles.

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.

C

Accommodation of differences between nominal and as-built geometry of both the piles and the superstructure. The most serious of these is generally believed to be the position of the piles, but in practice this is not always so - errors in girder geometry and alignment do sometimes occur.

C

Acceptance of a relatively large range of level in pile heads, that can be expected after trimming to length following driving.

C

Construction is unaffected by local damage to pile heads during driving.

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.

In-situ reinforced concrete deck slab

Precast deck beams Concrete pile capping beam

High Modulus Piles

Figure 5.2

Concrete beam ed on High Modulus Piles

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.

5.1.2 Steel column-piles As mentioned in Section 2, frame abutment bridges can also be built using individual bearing piles, ing each deck beam, rather than a retaining wall. These column-pile s are subject to much less resistance from the soil than walls, and they are also more flexible to the lateral displacement due to thermal and braking loads. Consequently, they attract less induced moment from the deck. P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

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 t


Road construction

Deck beam

Construction t

Concrete cross head beam /endscreen wall Select granular fill

Tubular pile Embankment fill

Figure 5.3

Built-in connection for tubular column-pile

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.


24

15 September 2003


5.2

Bankseat integral bridges

In a non-integral bridge, a bankseat abutment is normally essentially a simple L-shaped in situ reinforced concrete construction. The forces that it carries are the vertical reactions from the deck and the modest pressure from the retained pavement formation. If the soil strength is inadequate, bankseats may be founded on piles; the pile group can include raker piles to resist the longitudinal loads from the deck (when there are fixed bearings). Only vertical piles are required otherwise. In an integral bridge, the wall must be cast around the ends of the deck beams and the whole bankseat is then constrained to displace and rotate with the end of the deck.

5.2.1 Bankseat with spread footing


An integral bankseat detail that is ed 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 t

Deck beam

Select granular fill

Movement by sliding and rocking

Figure 5.4

Integral bankseat - spread footing

The deck beams will have to be ed at a construction t in the endwall and torsionally restrained during construction until they are cast into the end wall.

5.2.2 Bankseat with piled A piled bankseat can be used where a spread footing has inadequate bearing strength. Normal detailing of a pile cap will result in the piles of an integral bankseat bridge being subject to the displacement and rotation that occurs at the ends of the bridge. To avoid introducing additional restraint, and thus additional forces in the piles, only a single row of vertical piles should be used, as shown in Figure 5.5. This is the arrangement commonly found in the USA, where a single row of steel H-piles is used. Experience shows that the lateral resistance developed by such bearing piles under bridge deck thermal movements is negligible. P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

25

15 September 2003


The pile cap must be sufficiently deep to permit a practical transfer of vertical load from the pile. Asphaltic plug t

Deck beam

Select granular fill Reinforced concrete bankseat capping beam /endscreen wall

H pile

Figure 5.5

Integral bankseat, single row piled foundation


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

Pinned integral bridges

The provision of full moment continuity in frame abutment bridges necessitates the construction of a complex connection between deck and s, 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 deg 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 ing structure and transmit vertical and horizontal forces without relative displacement.

5.3.1 Pinned integral sheet pile retaining wall In the majority of design situations it is envisaged that it will be difficult to justify fully moment continuous connections between decks and substructures. Nonmoment inducing (‘pinned’) connections at the s will be the most attractive solution because they avoid complex design, are cheap to fabricate and aid practical and trouble-free fit up during installation. Provided that attention is given in design to durable details, it is possible that pin or knuckle bearings can be made sufficiently robust that they need never be removed over the full 120 years design life and hence provision for replacement can be avoided. This will permit economies to be made in design, construction and maintenance. For short bridges, the loads on the retaining wall will be quite modest and a simple sheet pile wall will suffice. A capping beam will be needed to spread the load from deck beams over the piles and to accommodate construction tolerances, but it can be relatively modest in size. In such a situation an effective pin can be created by


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 t

Reinforced concrete end screen wall

Deck beam

Holding down bolts set in grouted pockets Studs for load transfer


Reinforced concrete capping beam

Sheet pile or High Modulus Pile

Figure 5.6

Pinned-integral bridge, sheet pile retaining wall with bearings

5.3.2 Pinned integral High Modulus Pile retaining wall For larger spans, a retaining wall will need to use High Modulus Piles. In that case a more substantial pinned bearing is needed. A typical configuration is similar to that shown in Figure 5.6. During construction, the bearing lower plate (or at least its holding down bolts) can be positioned to suit before the concrete of the capping beam is cast. Further description of these simple bearings is given in Section 5.6.

5.3.3 Pinned integral on tubular column-piles Provision of a pinned bearing on top of a tubular column-pile can be achieved in numerous ways. Placing the bearing directly onto the top of the tubular pile-column is the simplest way. Figure 5.7 shows a section through a possible connection. A level surface at the desired elevation is best achieved by an in situ concrete filling to the column-pile. If the steel tubular column-pile is to be filled with concrete in any case for crash resistance, then this connection method will be the most efficient and suitable. This will provide a similar flexibility in positioning for the holding down bolts for bearings as that used traditionally for deck beams landed on r.c. abutments. Load transfer to the pile can be achieved using internal welded shear studs or weld bead fabricated and tested in the shop before delivery to site.


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 t

Reinforced concrete end screen wall

Deck beam

Waterproof membrane Holding down bolts set in grouted pockets Studs for load transfer


Embankment side slope Steel tubular column

Figure 5.7

5.4

Section through bearing connection detail for a tubular column-pile

tless deck configuration

A tless 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 t

Granular fill Sliding bearing

Figure 5.8

Flexible seal

tless deck bridge, spread footing bankseat


28

15 September 2003


As explained in Section 2.4, there is a horizontal movement interface between the underside of the endwall and the bankseat. This gap may either be open or filled with a non-structural filler. A flexible seal will be required across the end of this gap to prevent ingress of soil particles and surface water leakage. Such a seal is likely to be durable for a long period, since it is not directly exposed to the weather nor the rigours of the road environment and does not sustain any local loading. Asphaltic plug t

Granular fill Sliding bearing

Flexible seal


H piles

Figure 5.9

5.5

tless deck bridge, steel piled bankseat

Intermediate s

5.5.1 Tubular column-piles for intermediate piers Compliant intermediate s for multi-span beams can be effectively achieved by the use of steel column-piles. When they are concrete filled they offer superior vehicle collision resistance to that of concrete columns. Various arrangements for intermediate column-piles can be used and two and four column configurations are shown in Figures 5.10 and 5.11. A crosshead beam could be a practical structural arrangement which permits correction of any out-of-level or lack of verticality in the piling after driving and s individual pile columns to produce one effective intermediate structure.


29

15 September 2003


Transverse r.c. crosshead beam

Tubular steel column-piles eg. 965Ø 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 Ø x 40 mm w.t. steel column-piles

Figure 5.11 Four column-pile intermediate pier with steel crosshead beam


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


Cross head beam

Tubular pile

Figure 5.12 Tubular column-pile pier with pile capping beam and pinned bearing Moment connections between deck beams and intermediate s 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

Non-moment inducing (‘pinned’) connections at the s avoid complex design, are cheap to fabricate and aid practical and trouble-free fit up during installation. Provided that attention is given in design to durable details, it is possible that such bearings can be made sufficiently robust that they need never be removed over the full 120 years design life and hence provision for replacement can be avoided. This will permit economies to be made in design, construction and maintenance of pinned-integral bridges. Rocker and knuckle bearings are less common today than they used to be, having been replaced by proprietary elastomeric bearings (either steel/elastomer sandwich or elastomeric pot bearing). Nevertheless, they provide simple pinned connections that are easy to design. They are easily fabricated from thick steel plate to suit the particular load requirement, which gives a very durable detail and cuts out the cost of expensive high specification bearings. Little maintenance is required, other than P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

31

15 September 2003


occasional inspection, and even if a knuckle were to seize as a result of neglect, negligible moment is transmitted as a consequence. Such seizing is only likely on shorter bridges since on longer bridges where the deck thermal movements are significant, the t will be forced to work regularly.

5.6.1 Studies using pinned bearings Studies by the SCI(1) have examined the feasibility of using pinned bearings for tubular column-piles, and representative cases have been analysed for intermediate and end s for a symmetrical two-span bridge with each span of 40 m.


At the intermediate of such a two-span overbridge carrying a typical minor road, the maximum vertical reaction at Ultimate Limit State will be of the order of 3 MN. Lateral forces from wind loads and skidding forces are fairly small (300 kN nominal skidding load for example). The greatest horizontal load will be that arising from collision loading on the substructure (in accordance with BD 60/94(23)), in which a column-pile will behave as a propped cantilever between an effective fixed position below ground and the bridge deck connection. Longitudinal displacements at the intermediate of such a two-span bridge were found to be very small, with rotations (in the plane of the girder) of less than ± 0.01 rad. In a transverse plane the bridge girders were effectively braced and there was little tendency to rotate about the longitudinal axis. The use of linear bearings (which effectively provide a torsional restraint to each girder) is therefore quite suitable. At the ends of such a bridge, the maximum vertical load is smaller, but the same lateral braking load must be catered for. If the columns are not protected and are within 4.5 m of the nearside lane edge, collision loads must also be allowed for according to BD 60/94. Rotational displacements in the girder plane will be less than at intermediate s. Maximum thermal longitudinal displacements of about 20 mm can be experienced, but since the steel column-piles are quite flexible, this will only give rise to small forces. Typical configurations of rocker and knuckle bearings are shown in Figure 5.13.

Rocker bearing

Knuckle bearing

Figure 5.13 Simple pinned rocker and knuckle bearings

5.6.2 Rocker bearing To resist the magnitude of reaction at an intermediate (around 3 MN or less), a simple fabricated linear rocker approximately 600 mm long, comprising a cylindrical upper plate and a flat lower plate, is adequate. A suggested configuration is shown on top of a thick capping plate to a pile-column in Figure 5.14. The key features of this bearing are described below.


32

15 September 2003



Figure 5.14 A simple pinned rocker bearing The bearing can be fabricated from ordinary structural steel, grade S355, although there may be advantages in using higher grades for the rocker plates. A cylindrical surface of radius about 250 mm, or greater, can be machined on the upper rocker plate. Upper and lower rocker plates are about 600 mm long and about 100 to 120 mm wide. Friction cannot be relied upon to resist horizontal forces, so shear keys or dowels must be used. In view of the relatively high shear force arising from collision loads, several ‘shear dowels’ will be required. These carry all the shear force and it would help in minimising their size if higher strength steel is used, but this is not essential. Dowels of about 35 to 40 mm diameter are suitable. Since there are no ‘moving parts’, such bearings are durable and easily maintained. Connection of the rocker bearing to the underside of the deck beam and to the top of a column-pile is via plates which are welded to the rocker plates (only the lower plate is shown in Figure 5.14). The rocker bearing plates can either be bolted or welded to the deck beam and set in concrete with anchor bolts at the top of columns or abutment walls.

5.6.3 Knuckle bearings In view of the combination of large horizontal force with, possibly, only a small vertical force, knuckle bearings can be used for pinned connections. Such bearings have been used on older and heavier bridges, and as bearings for arches. The loads considered here (3 MN) are fairly modest for a knuckle bearing, and could easily be accommodated by a pin of less than 100 mm diameter. No special measures are needed to cater for horizontal forces. However, the type of knuckle bearing shown in Figure 5.13 is more complex to fabricate, and thus more expensive, than a rocker bearing, and it does act by sliding (around the cylindrical surface), which would require more frequent inspection and attention during maintenance than the pinned rocker.

5.7

Skew bridges

Previous discussion and illustrations of the behaviour of integral bridges have considered the bridge in elevation, or in the plane of one of the main beams. However, very many bridges are skew in plan, some with a very large skew angle.


33

15 September 2003


Where an integral bridge is ed 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.


Figure 5.15 Rotation of a skew deck on wall-type abutments In BD 57/95(8), the Highways Agency has recognised the problems involved in predicting the complex behaviour of skew bridges and (in principle) only requires bridges of skew up to 30E to be designed as integral. However, if an integral bridge is ed on pile-columns, the rotational effect is eliminated. Similarly, skew bankseat integral bridges will not rotate significantly though they may slide parallel to the face of the endscreen wall. tless deck bridges can also be constructed with skew. The Highways Agency limitation therefore should be regarded as a warning, rather than an absolute limitation.


34

15 September 2003


6 STEEL SECTIONS FOR INTEGRAL BRIDGE PIERS AND ABUTMENTS This section reviews the various steel profiles and sections that can be used in the sub-structures of integral and tless deck bridges. Sections for continuous walls and for individual s are discussed.

6.1

Continuous wall steel pile sections


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.

6.1.1 Frodingham sections (Z profiles) Frodingham sheet piles are proprietary products that are manufactured by British Steel. Each pile interlocks with the adjacent pile during driving to create a continuous sheet pile wall structure. For Frodingham sections (Z profiles), the interlocks are between adjacent piles at their outer faces. A typical cross-section detail is shown in Figure 6.1.

Figure 6.1

A Frodingham sheet pile

Frodingham sheet piles have advantages in a marine environment as the interlocks are generally more watertight than Larssen piles. Also Frodingham sections tend to be preferred for cantilever walls, as horizontal deflections towards the excavation tend to be lower. The section modulus for the Frodingham profiles ranges from 688 cm3/m for the 1BXN section to 3168 cm3/m for the No. 5 section. The complete range of Frodingham sections and their properties are listed in British Steel’s Piling Handbook(24) which can be obtained from British Steel Sections, Plates and Commercial Steels, Scunthorpe.


35

15 September 2003


6.1.2 Larssen sections (U profiles) Larssen sections are also proprietary products manufactured by British Steel. For Larssen profiles the interlocks are positioned on the centre line of the wall (see Figure 6.2). The piles depend upon friction at the interlocks, to develop the full modulus of the combined section. As the interlocks are close fitting, this form of construction causes a considerable amount of friction to be developed. In most cases, the friction at the interlocks is achieved by shear being generated by surface irregularities, rusting, lack of initial straightness and soil particle migration into the interlocks during driving. This interlocking enables the full structural capacity of the combined section to be developed on most walls (see British Steel’s Piling Handbook(24), pages 1/7 and 5/15). If there is concern about achieving full structural capacity for a very high wall where it could be critical then the interlocks can be welded. Crimped or welded pairs can be delivered to site. Further advice will be available in Eurocode 3: Part 5(25), when it is published.


Figure 6.2

A Larssen sheet pile

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 Steel’s Piling Handbook. Larssen 6 piles were used in the retaining walls for the frame abutment integral bridge at the Bridgend-By-, see Section 4.2.

6.1.3 High Modulus Piles A High Modulus Pile is a proprietary product manufactured by British Steel Piling which comprises Frodingham sheet piles welded to the outer surface of the flange of a Universal Beam. A typical section is shown in Figure 6.3. Each pile is driven clutched to an adjacent pile to form an interlocking sheet pile wall acting compositely with the Universal Beams. The resulting profile provides a wall of identical repeating units. High modulus piles are produced in a range of sizes such that the calculated structural capacity can be very closely matched, thus providing maximum economy.


36

15 September 2003


Intermittant weld at interlock 75mm weld at 600mm crs with 300mm closure weld at ends 966mm for 3N & 4N 850mm for 5 Width of one combined pile 283mm for 3N 330mm for 4N 311mm for 5

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

Continuous fillet weld x

Universal beam

Figure 6.3

High Modulus Piles


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 Steel’s 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 ing 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).


37

15 September 2003


Weld

Weld Weld

Figure 6.4

Types of box piles

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 ing 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.


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

5

0.2

0

0

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

British Steel available range of linepipe

H-Piles

Steel H-piles have been used extensively in bearing pile s for bridges in the USA and the UK. A recent example is their use on the A47 Walpole to Tilney End By- 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 Steel’s 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 ed can either be greater, or if soil conditions permit, the driven steel piles can be shorter to a given structural load. Further information is given in the Steel bearing piles guide(12). y

x

x

y

Figure 6.6

A typical H-Pile section


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


The accuracy of pile installation is of the utmost importance to minimise on site connection preparatory work at the bridge deck to pile head interface. However, tolerances must be allowed in the positioning of driven steel piles, and it will be necessary to consider lack of fit forces and local moments in the design of the connection details. Information on tolerances that are achievable using commonly available pile driving equipment and methods are quoted in the ICE publication Specification for piling(26) and specifications issued by the Federation of Piling Specialists(27), the Eurocode TC 288 WG4(28) Execution of sheet pile wall structures and the TESPA Installation of steel sheet piles booklet (29), and the SCI publication Design guide for steel sheet piled bridge abutments(30). However, only the least onerous levels of tolerance are given in those publications. Following discussions with of the Federation of Piling Specialists (FPS), a summary of the most accurate or best alignments that can be expected using specialist equipment, skilled personnel and careful planning of procedures in respect of site conditions was drawn up and this is presented in Table 6.1. Table 6.1

Tolerances that can be met in driving steel piles

Type of pile and method of driving Deviation normal to the wall centre line at pile head

For pitch and drive method or over water

For drive method

±25 to 75 mm Dependant on equipment used.

Finished level deviation from a specified level of pile head, after trim

±2 mm

of pile toe

±120 mm

Deviation from specified inclination measured over the top 1 m of wall Normal to line of piles Along line of piles

±1.5%

±1%

±1%

±0.5%

Installation planning is essential to allow for any inhomogeneity in the soils. Accuracy of alignment will also be affected by pile stiffness, the driving equipment and the competence of the workforce. Conventional pile guide frames (formed of stiff universal beams with the webs horizontal) ensure good alignment for sheet piles in driving methods. It is also important that piles be finished to level with minimal tolerances and that any damage to the top of the pile is cut off. Experience of steel sheet pile driving shows that over length piles should be supplied to allow for on-site changes in soil conditions between boreholes. This overdrive allowance is normally +10%, with spare piles being provided to replace any that are severely


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

Environmental factors associated with driven piles


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 . If the designer has little experience of assessing driveability, assistance can be sought from a specialist piling company recommended by the FPS.


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.


A further computer program derivative of the wave equation is CAPWAP which permits the analysis of recorded stress waves in steel piles during driving. Such a PDA (‘Pile Driving Analysis’) enables the interpretation of soil resistance by curve fitting techniques. The force-time recording from strain gauges on the pile is compared to the force-time curve produced by integration of the simultaneous acceleration-time recording. The latter is adjusted by varying the soil resistance model on the pile shaft and base, thereby gaining an understanding of both its magnitude and distribution down the pile. By comparison with static pile load tests on the same pile, an extensive database has now been assembled by many practitioners throughout the world (see the Proceedings of the International Conferences on the Application of Stress Wave Theory on Piles (1980, 1984, 1988, 1992, 1996)(34). As a result, the static load resistance of piles can be predicted to an accuracy of +/- 10% in granular soils and rocks. Care must be exercised with the use of the method in clays however because of the time delay in shaft friction recovery after driving.

6.8

Corrosion allowances

The means for countering the effect of corrosion of steel piles are well developed. Guidance is presented in British Steel’s Piling Handbook and in the Highways Agency’s 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.

6.8.1 Corrosion and protection of steel piles According to BD 42(35), steel piles are to be designed “with sacrificial thicknesses applied to each surface, depending on the exposure conditions, to provide a design life of 120 years”. Table 6.2 summarises the required sacrificial thicknesses for each surface for different exposure conditions, which are based on the advice given in BS 8002(36) Clause 4.4.4.4.3. The reduced (corroded) section properties can either be obtained by calculation or can be obtained from British Steel’s Piling Handbook. Table 6.2

Sacrificial thicknesses for piling according to BD 42


Exposure zone

Sacrificial thickness (for one side of the pile only) (mm)

Atmospheric

4

Continuous immersion in water or effluent

4

In with natural soil

2

Splash and alternating wet/dry conditions

9

The procedure to determine corrosion allowances for a particular situation in accordance with BD 42 can be presented in the form of a flowchart - see Figure 6.7. It should be noted that the corrosion allowances in BD 42 all apply to unprotected steel piles. Although it is generally cost effective to provide the sacrificial steel thickness, consideration can be given to the following alternative corrosion protection options: C

Protective coatings, particularly in the exposed section of the pile.

C

The use of copper-bearing steel (effective against atmospheric corrosion).

C

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.

C

Cathodic protection in soil below the water table or in a for marine environment

C

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 Steel’s 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


43

15 September 2003


sacrificial thickness. No guidance on the use of cladding or coating to reduce the corrosion allowance is available in BD 42, although it is understood that this is currently being reviewed for a future revision of that standard. British Steel has undertaken significant research and development into corrosion of steel and corrosion protection and further advice can be obtained by reference to The corrosion and protection of steel piling in temperate climates(39) and The prevention of corrosion on structural steelwork(40).

6.8.2 Corrosion in fill or industrial soils Bridges are sometimes constructed in areas of recent fill or ‘industrial soils’. Corrosion protection of the steel in with the fill material may be required. and this can be assessed by testing the material for pH and resistivity.


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 CIRIA’s series of reports on contaminated land ( 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.


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


Yes Sacrificial thickness 4 mm

Natural Undisturbed soil ?

Yes


No Classify soil from Tables 1, 2 and 3 of BD 42/94

Non-aggressive soil ?

Yes


No

Very aggressive soil ?

No


Yes Seek specialist advice

Figure 6.7

Determination of corrosion allowance


45

15 September 2003


7 DESIGN BASIS 7.1

General principles

Limit state design should be adopted as far as possible for integral and tless 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 of the interaction between structural and geotechnical elements (such an analysis is not yet covered in any design code).

7.2

Design standards

7.2.1 National and European 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


46

15 September 2003


strength is obtained from the ‘representative strength’/1.2, where 1.2 is the mobilisation factor. No analysis at the ultimate limit state is then required by BS 8002 because it assumes that the forces acting on the retaining wall at the serviceability limit state are greater than those at the ultimate limit state. According to the research by Potts and Fourie(46), which is quoted in the ISE publication Soil-Structure interaction - the real behaviour of structures(47), the earth pressures generated are related to wall movement, and much more precision is required in the definition of serviceability limit state before the ultimate limit state can be dismissed as in BS 8002.


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.


47

15 September 2003


Table 7.1

Partial factors to be applied to actions and ground properties, according to DD ENV 1997-1 Actions Permanent

Case

Factor on Ground Properties Variable tan N

c’

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

Loss of static equilibrium; strength of structural material or ground insignificant Failure of structure or structural elements, including those of the footing, piles, basement walls etc., governed by strength of structural material Failure in the ground Angle of shearing resistance Undrained shear strength


7.2.2 Highway Agency Standards The Highways Agency sets out its requirements and advice for the design and construction of structures in Standards (BD’s) and Advice Notes (BA’s). The BD documents contain mandatory requirements that are effectively the ‘building regulations for bridges in the UK’. Some of the BD documents are used to invoke BSI codes of practice as mandatory design rules, subject to a few amendments to suit the HA’s particular requirements. The BA documents give background information, guidance on the application of the related BD and generally recognised rules that satisfy the requirements of the BD. The BAs are not intended to be mandatory. As explained in Section 3.2, the requirement that integral bridge construction be considered is given in BD 57/95. No Standard has yet been issued relating to the design of such bridges, but guidance is given in BA 42/96. However, it must be noted that it does not provide comprehensive design rules. It only sets out advice relating to design and construction with a view to achieving greater consistency in approach between designers. The Highways Agency has confirmed in a Seminar on Integral Bridges, held at the Institution of Structural Engineers, London, on 23 January 1996 that BA 42/96 is not mandatory. Where piled abutments and/or piers are specified, BD 32 Piled Foundations(49) sets out the design and construction requirements, with BA 25(50) providing guidance on its use and interpretation. It may be noted that determination of design load effects are specified in BD 37(51), which introduced a modified version of BS 5400: Part 2. This includes the specification of partial load factors on earth pressures and on the restraint of thermal movement. Such factors may not be appropriate to integral bridge construction; see further comment in Section 12.4. Where bridge s are located within 4.5 m of the edge of a carriageway, they are required to be designed to withstand vehicle collision loads. The magnitude and application of this loading is given in BD 60/94.


48

15 September 2003


7.2.3 Other standards The design of tubular piles is not covered by a UK or CEN Standard, but is comprehensively covered by API RP2A(22). This is an American Code of Practice which is used by the worldwide oil and gas industry to design fixed offshore installations. Using the API RP2A document as a basis, ISO 13819-2(52) is under development for international use. These documents are based on a limit state philosophy using partial factors.


7.2.4 Material properties for steel piles Steel sheet piling, including sections for box piles, are produced in accordance with BS EN 10248(38); grades S270GP and S355GP are commonly used. Universal beams (for High Modulus Piles) and other plates and sections are produced in accordance with BS EN 10025(53); grades S275 and S355 are commonly used. Whilst the S355 strength designations are the same, the difference between S270 and S275 should be noted. Traditionally, impact toughness has not been considered a requirement for piling in the ground and so the grades S270GP and S355GP have no toughness testing requirement (although the Standard provides an ‘Option’ to require toughness testing). But, for steelwork designed to BS 5400: Part 3, notch toughness is required, although the ‘service temperature’ for steel in with the ground may not be as low as the minimum effective temperature of the bridge deck. In the absence of specific guidance from the Highways Agency, it would seem not unreasonable to specify a toughness requirement equivalent to J0 quality (27J at 0EC) for the sheet piling and grade S275J0 or S355J0 for the UB’s; as mentioned, the sheet piling material can be tested on request, and it is likely to meet the requirements of J0 quality without any special measures. Tubular steel piles are produced in accordance with API 5L(54) rather than a BSI or CEN Standard. The specified minimum yield, according to that Standard, may be used for design. Values of specified minimum yield for steel grades to API 5L are given in Table 7.2. There is a toughness requirement set at a Charpy value of 35 for plate thicknesses of between 20 to 100 mm at !40E C for the weldability of these steels in offshore conditions; this is in excess of that required for onshore piling. Other useful information on the specification and use of structural steel tubulars is contained in Appendix 21 of the Offshore Installations: Guidance on the design, construction and certification published by the HSE(55). Table 7.2 API Grade

Yield stress for cold formed line pipe Specified minimum yield stress MPa

X46

317

X52

358

X56

386

X60

413

X65

448

X70

482

X80

551


49

15 September 2003


7.3

Limit state design

Limit state design considers two principal limit states: C

Ultimate limit state

-

at which collapse or other form of failure occurs

C

Serviceability limit state

-

the states prior to collapse beyond which specified serviceability criteria are no longer met.

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 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)

γ f3 γ m

* and S = ( effects of γ

fL

× design loads)


Whereas in Part 4 they are expressed as:

R* =

function(characteristic strength)

γm

and

S* = γ

f3

( effects of γ

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 Steel’s 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

Loading effects on foundations

Design load effects on the foundations of an integral bridge arise from a variety of sources: C

Soil pressures.

C

Ground water and free water pressures.

C

Seepage forces.

C

Embankment surcharge loads.

C

Ground movements.

C

Bridge structure self weight.

C

Traffic loads.

C

Braking loads.

C

Temperature changes in the deck.


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

Observational Method for foundation design

The Observational Method involves making the best estimate of a geotechnical behaviour, in conjunction with formulation of contingency plans for additional measures to be taken if the actual behaviour exceeds predictions by an unacceptable margin. In the construction industry, increasing emphasis is being placed on the value of the Observational Method (see Peck(56)) whereby immediate from instrumentation monitoring of retaining wall behaviour is used to modify the design and construction procedures according to a pre-determined plan. In geotechnical engineering the current state of the art is such that predictions of wall and pile displacements are subject to a considerable degree of uncertainty. Among the reasons for this is the difficulty in predicting the soil response to structural loading from a limited number of tests on soil samples, and the general lack of monitoring of real highway structures for correlation. Where a Serviceability Limit State deflection is judged by the designer to be really necessary and is specified for a retaining wall, say where protecting an adjacent building from the effects of the bridge works, then the Observational Method may be the only reliable method to permit control of the works. The Observational Method is recommended in the ENV-1997-1 Eurocode 7(48). It states that if this method is to be used it is imperative that before start of construction the following requirements are met: C

Acceptable limits of behaviour are established.

C

The actual behaviour lie within the acceptable limits.


51

15 September 2003


C

A monitoring plan is set up which shows that the actual behaviour lies within the acceptable limits.

C

A contingency plan is available if the actual behaviour is outside 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.


The Highways Agency is considering the implications of the use of the Observational Method for steel and concrete piled retaining structures and is funding research in this area. The TRL have started this process for concrete retaining walls(57) but not yet for sheet pile walls. Various studies on the use of the Observational Method for highway structures have been or are in the process of being published by the Transport Research Laboratory(58) and others(59). Numerous highway structures have been built using this method, for example the A406 under at Neasden and the Limehouse Link Tunnel. Operational experience of the use of instrumentation data to temporary works design has been documented on the Aldershot Road Under, Guildford(60). Monitoring the behaviour of structures as part of the observational method provides a reliable means of evaluating the validity of current design methods, from comparisons of the measured to predicted behaviour. In the design of an integral bridge, which requires an appreciation of the behaviour of soils in situations not previously encountered, the Observation Method may prove a valuable design tool.

7.6

Design report

Although not a requirement of the Highways Agency (but a requirement of ENV 1997-1 Eurocode 7(48)) the assumptions, data, calculations and results of the verification of safety and serviceability of bridge substructures should be collated and presented clearly in a design report. The report should include, but not be limited to, the following: C

An examination and description of the site.

C

Interpretation of the results of ground investigation.

C

Analysis of the results of site monitoring of ground water conditions.

C

A list of the geotechnical design assumptions.

C

Any anticipated geotechnical problems and statements on the contingency actions to be taken.

C

Statements on geotechnical requirements for the design and construction of the sheet pile retaining wall including testing, inspection and maintenance requirements.

C

Stability analyses of the site and calculations for the structural design of the steel sheet pile wall.


52

15 September 2003


7.7

Design for fatigue

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).

C

120 annual cycles between summer and winter temperatures (estimated from TRL 696).

C

365 × 120 daily cycles between day and night temperatures (estimated from TRL 696).

which is summarised in Table 7.3. Table 7.3

Bridge temperature cycling

Time Period

Extreme (120 year) Annual Daily

Cycles

Temperature changes Effective EC

Difference EC

1

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 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 tless 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 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 s 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.


54

15 September 2003


Site investigation

Select initial concept

Design deck Evaluate benefits of continuity deck/

8.2.1

8.2.2

8.2.3

Select soil properties 8.2.4

Select retaining wall, column-pile or bankseat configuration


Design abutment or bankseat structure

Intermediate s?

Design for reactions from deck Evaluate benefits of moment continuity

Yes

No

Include continuity in interaction Carry out interaction analysis (deck + s)

Check adequacy of deck and s

Revise deck and/or s

Strengths OK ?

No

Yes

Design details, including connections deck/s

Figure 8.1

Overall design sequence for an integral bridge


55

15 September 2003


8.2

Preliminary stages

8.2.1 Site investigation A good knowledge of the site conditions and relevant soil data are essential in deg a safe and efficient retaining wall bridge abutment. The designer must be confident of the data; further boreholes may be needed to better delineate the soil boundaries over the site. A geotechnical specialist should be involved in specifying the site investigation techniques to be used. Refer to Section 9.2 for guidance on site investigations and Section 9.3 for an explanation of the selection and evaluation of soil parameters. The soil data to be obtained will include bulk densities, effective stress strength parameters (c’, N’), deformation and stiffness parameters, and ground water conditions. BS 8002(36) provides some useful further information on representative values for different types of soil.


8.2.2 Select initial concept The conceptual design of an integral bridge structure may be performed using the guidance given in Sections 5, 6, 7, and 8. Section 5 explains the different forms of and connections between bridge deck and the substructure; both moment resisting and moment-free connections are considered. Section 6 describes steel pile sections that can be used for bridge substructures. Pile driveability and construction sequence are very important aspects that need to be considered in conceptual design because they affect the practicality, loading, and cost of integral bridges. Decisions will need to be taken at the conceptual design stage on the type and configuration of integral bridge that is likely to be the most suitable for the site; for the client requirements; and for the lowest whole life cost before proceeding with more detailed design. Where construction cost is the overriding criterion, several alternative types of bridge may need to be evaluated in more detail in order to arrive at reasonably detailed design drawings that can be priced more reliably by contractors. The construction sequence generally affects the load combinations that govern integral bridge structure design and hence it needs to be decided before commencing detailed design, particularly where soil-structure interaction will be involved. The cost of building the structure will only be part of the comparison because steel substructure bridges offer faster construction than many conventional types and this can significantly reduce the overall construction cost. The saving in construction time may be a very useful advantage to the client, especially where lane rental charges are to be applied and minimum disruption to traffic on bridge replacement schemes is an important factor.

8.2.3 Initial design of deck The initial design of a bridge deck (selection of girder size, flange dimensions, spacing, etc.) will be the same as for a non-integral bridge except that there are particular regions that require more attention; these include the bottom flanges in the region of intermediate s and at the abutments of frame abutment integral bridges. At these locations, hogging moments will be present and the bottom flanges will be in compression; longitudinal compression due to restraint of thermal P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

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 s 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 s. 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 s, 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 s condition is not considered. No detailed advice is given in this publication on bridge deck design; guidance is available from other SCI publications(65).


8.2.4 Select soil properties Soil properties relevant to the type of foundation should be extracted from the site investigation report. Discussions may need to be held with geotechnical specialists to achieve this (see Section 9).

8.3

Design of embedded retaining wall abutments

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.

8.3.1 Wall stability and depth of embedment An analysis is required to check the overall stability of the embedded retaining wall against overturning, and to calculate the required depth of embedment. The most appropriate analysis is one based on a limit equilibrium method; this is explained in Section 10. In that section the Factor on Strength method is recommended and the results checked using the Burland-Potts method. The recommended partial and/or lumped factors of safety to be used are given in Section 10. Factored highway loadings are applied at the relevant points and at the appropriate levels, in combination with unfactored permanent loads and any other loads. Partial load factors are not applied to earth pressure because an overall lumped factor of safety approach is used in this part of the design, applying the factor to the load effects at the end of the analysis for design of the wall. Live load surcharges are only applied to the retained side of the wall. See Section 7.4 for a summary of loads to be considered.


57

15 September 2003


Perform wall stability analysis to determine depth of embedment & initial design moment for wall section

Select a pile section

8.3.2

Determine depth of embedment to resist vertical loads

Perform pile driveability analysis


8.3.1

8.3.3

8.3.4

Choose frame or pinned abutment

Frame abutment

Pinned abutment

Determine retaining wall boundary conditions due to axial stiffness & rotational stiffness of deck

Determine effective prop stiffness due to deck 8.3.6

Carry out retaining wall analysis (FREW, WALLAP, etc) Determine bending moments and forces 8.3.7

Carry out propped cantilever wall analysis (FREW, WALLAP, REWARD, BA42 method) Determine bending moments and forces

Check wall & deck for moments and forces at ULS and SLS

Figure 8.2

Design sequence for embedded retaining wall abutments


58

15 September 2003


8.3.2 Selection of sheet pile section size A preliminary pile section for the abutment retaining wall should be chosen from the types of continuous steel sheet pile sections that are commercially available. These include sheet piles (Frodingham and Larssen Profiles), High Modulus Piles and box pile combinations. In the selection of section size the following requirements should be considered: Structural capacity The initial selection of pile size should be made on the basis of moment resistance to sustain the (factored) maximum moment determined by the wall stability analysis. For advice on the basics for calculation of steel section resistance, see Section 7.3. Corrosion The maximum appropriate corrosion allowance for the exposure conditions as specified in BD 42/94 should be deducted before determining the structural capacity and selecting a section at this stage of design (see advice given in Section 6.8).


8.3.3 Vertical resistance of sheet pile walls The length of the pile section chosen for its bending resistance must also be checked for its adequacy to carry axial loads and then the depth of embedment increased where necessary. A method of predicting the load resistance of driven steel piles is given in Section 13. If the depth of embedment is inadequate to resist the vertical loads from the bridge deck, then an alternative is to use a larger pile section size.

8.3.4 Pile driveability and dynamic analysis The designer should check that the chosen steel pile can be driven to the required penetration. This can be done either by using pile driving formulae or by using wave equation analysis. See Section 6.7 for guidance. The objectives of the pile driveability analysis for the selected pile section size are:

C

To check that the pile can be driven to the required penetration.

C

To check that the pile will not be overstressed during driving.

8.3.5 Choice of frame or pinned abutment A choice needs to be made between either a pinned or framed abutment configuration at this stage because the type of deck/substructure connection affects the type of analysis required. The benefits of selecting a frame abutment configuration should have been evaluated in the preliminary stages and a clear plan should have emerged to enable the selection at this juncture.

8.3.6 Boundary conditions Software is not yet available that will model accurately the soil behaviour in an integral bridge at the same time as the structural behaviour of the deck as a whole. Consequently, separate analysis must be made and the interaction between then evaluated.


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 s. 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 s 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.

8.3.7 Bending moments and forces acting on the abutment Frame abutment For a framed abutment, a soil-structure interaction analysis will be required to solve the strain related earth pressure problems. Suitable numerical analysis programs are FREW and WALLAP; both are widely available in the design industry. See Section 11. Load cases will be run for all the situations during construction and in service. The effect of thermal expansion of the deck can be modelled directly using numerical analysis programs because they are based on the compatibility of stress and strain. The detailed modelling technique will vary between programs, but the principle is to induce the thermal displacement into the soil model at the deck level. For braking load cases, the frame abutment is analysed as a sway frame, and should be taken of the sway response of the deck that creates an asymmetry in abutment load. The most critical load combination is again difficult to prejudge


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.

Bending moments due to construction tolerances and deformations Additional bending moments are caused by out-of-vertical and positional tolerances in construction and when lateral load displacement occurs, and these need to be included in the total effects. Typical misalignments due to tolerance in construction are given in Table 6.1 and horizontal displacements are obtained from the soilstructure interaction analysis.

8.3.8 Structural adequacy of the abutment The bridge abutment has to resist all combinations of loads at both ULS and SLS (see Section 7). Steel pile sections are to be checked for adequacy in strength both for static loads and dynamic pile driving forces (see Sections 6 and 7). Where


61

15 September 2003


appropriate, adequacy of section is to be considered for uncorroded and corroded cases (see Section 6.8). For static loading, adequacy of section has to satisfy the requirements of bending, axial and shear including the interaction between bending plus shear force and bending plus axial force interaction. BS 5400(42) does not provide a check for bending, shear force and axial force interaction, therefore it is necessary for the engineer to apply a rational interaction check which will give the most conservative outcome (see comment in Section 7.3 on checking the resistance of the deck). Adequacy of structural strength of the abutment at SLS is made by checking the wall displacement output from the soil-interaction analysis (see Section 7.3). If the data indicates that structures or services may be affected, it will be necessary to perform a deformation analysis to determine ground movements. Wall displacements for both the construction and in-service stages are determined either by finite element methods (see Section 11) or by use of the Observational Method (see Section 7.5).

8.3.9 Detail design - connections and fatigue analysis


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.

T32 at 150 crs

8 rows of 3 No. 25mm dia. studs at 150 mm crs 8 No. T25 at 150 crs.

7 No. T40 at 200 6 No. Hoops

Construction ts

T40 at 150 crs

500 x 600 x 100 deep landing plinth

Neutral Axis of High Modulus Pile

Figure 8.3

Section through pile cap showing reinforcement details


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 t 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

Design of column-pile abutments and piers

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 s, is given in the same manner as the preceding Section 8.3 does for embedded walls.


8.4.1 Axial load resistance The axial load resistance of a tubular or box section bearing pile that is driven open ended is covered in the Steel bearing piles guide(12). Some of the advice is repeated here in Section 13. Pile driveability and dynamic analysis See Section 8.3.4.

8.4.2 Laterally loaded pile analysis For the pile size and configuration determined from axial load capacity requirements, the tubular pile abutment/pier has to be checked to confirm that it has adequate strength to resist the bending moments and forces that act on the steel piles due to lateral loads acting at the deck to pile connection. A check has also to be made on the earth pressures that may act due to adjacent embankment loading. Guidance on performing the analysis of laterally loaded bearing piles is given in Section 14. For further details on the types of analysis programs, refer to the CIRIA Report Design of laterally loaded bearing piles(67) and Steel bearing piles guide(12).


63

15 September 2003


Select a pile section for axial resistance and check driveability

8.4.1

Carry out lateral analysis of individual pile (thermal, braking and collision loads)

8.4.2

Check pile for combined stresses Adjust wall thickness and diameter

Frame abutment integral

Pinned integral


Impose deck rotations on pile calculate BMs in pile and connection

Check pile for combined stresses

Check pile length for toe fixity

Design pinned connection and bearings

8.4.3

8.4.3

8.4.3

Design pile cap moment connection

Check deck

Figure 8.4

Design sequence for column-pile abutments and piers

8.4.3 Pile/structure interaction Frame integral bridge The imposition of a moment connection at the head of the pile where it is embedded into the pile cap requires a moment/rotation diagram to be developed for the pile head. This can be obtained using lateral load pile analysis programs like ALP in the Ove Arup OASYS suite(68).


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- 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 es 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.

8.4.4 Design of pile cap The concrete pile capping beam or pile cap is reinforced to resist all the load effects for each load case combination, including crash resistance reactions where that is required.

8.4.5 Crash resistance analysis of tubular piers Where column-piles are close to the carriageway, BD 60/94(23) requires that they be designed for resistance to collision loads. The Standard specifies equivalent static loads, which can be used for overall design, but which are not suitable for local design of a hollow steel column. Further advice is given in Section 14.5.

8.5

Design of bankseat integral bridges

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).


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.


66

15 September 2003


9 SITE INVESTIGATION AND SOIL DATA FOR DESIGN This section provides the designer with an awareness of the soil and ground data that is required to perform an accurate design of an integral bridge. Accurate data is required about the site and the soil, therefore a knowledge of the methods and procedures that are to be performed by a geotechnical engineer to obtain this data must be understood by the designer. In some cases soil properties may not be initially available or previously determined values may be insufficient for design. Supplementary boreholes would then be required that are tailored to the particular requirements of integral bridge analysis.

9.1

Soil data required for design


The soil data that is required for pile design includes: C

Saturated and unsaturated bulk densities (unit weight).

C

Undrained and drained shear strength including angle of shearing resistance. and cohesivity.

C

Standard Penetration Test resistance N values for pile end bearing and shaft resistance.

C

Cone penetrometer resistance, qc .

Deformation and stiffness soil properties required include but are not limited to: C

Young’s modulus (E50 and initial tangent modulus).

C

Poisson’s ratio.

C

Coefficient of horizontal subgrade reaction.

C

Coefficient of subgrade reaction.

C

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.


67

15 September 2003


Before a method of construction is proposed and a design commenced, it is essential that site explorations are considered, and performed where required. Site investigations are performed to ascertain the character and variability of the strata underlying the site of the proposed retaining wall and adjacent to it. In particular, those properties that could affect the performance of the structure and the choice of the method of construction should be assessed thoroughly. Loads on bridge foundations and stability of walls are significantly influenced by ground water conditions. The ground investigation should therefore include the installation of standpipes to monitor the ground water table and the geotechnical report should interpret the seasonal and long term fluctuations that could be expected.


Although site investigations are important in enabling an accurate design to be produced, the extent and detail of investigation has to reflect the type of structure that is to be designed. ENV 1997-1(48) provides classification categories for geotechnical design requirements, which can be adopted in assessing the extent and detail that is required for a particular project. These are referred to as Categories 1, 2 and 3 (see Section 2.1 of ENV 1997-1). For some complex projects, site investigations will need to be performed during construction (see the Observational Method discussed in Section 7.5). Periodic ground inspections during construction enable actual conditions prevailing to be monitored and soil parameters modified as the design is progressed.

9.3

Selection and evaluation of soil parameters

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 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

Soil parameters for design of integral bridges

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 τ

100 Moderately conservative o (c'=10 kN/m 2 , φ ' = 21 )

50

0

Worst credible o (c'=0 , φ ' = 21 )

50

100

150

200

Effective stress (kN/m2 ) σ '


Figure 9.1

Representation of a representative (worst credible) value

Care must be taken in using quoted values or specifying soil properties to be measured as various other definitions exist. Characteristic values as defined in ENV 1997-1 are selected as a cautious estimate of the values affecting the occurrence of the limit state. Usually the extent of the ground governing the behaviour of a pile abutment or pier at a limit state is much larger than the extent of the zone in a soil and consequently the governing parameter is often a mean value over a certain surface or volume of the ground. In this case the characteristic value is a cautious estimate of this mean value. The CIRIA Report Design of retaining walls embedded in stiff clays(66) takes into the uncertainty in the selection of soil strength parameters for clays by allowing two different values to be defined. These two values are termed Moderately Conservative and Worst Credible and are defined below. “A moderately conservative value for a soil parameter is defined as a conservatively best estimated value”. It is the most commonly used in practice by experienced engineers. “A worst credible value is the worst value that the designer could realistically believe might occur”. In the case of a load or a geometric parameter, it is not the worst physically possible, but rather a value which is unlikely to be exceeded. For example, in the case of a soil strength parameter, the worst credible value appropriate to a retaining wall design would be a pessimistic value which is very unlikely to be lower.


69

15 September 2003


10

ABUTMENT WALLS - EMBEDDED WALL STABILITY

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 ing 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 conditions, whether for cantilever or propped walls, it is assumed that there is enough movement of the ground to allow full active and ive 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 Steel’s 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 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).

10.1 Cantilever and propped walls As a cantilever wall depends entirely on the of the penetrated ground for its stability, it cannot be in equilibrium without the toe being prevented from rotating. Embedded cantilever retaining walls are therefore designed using the fixed-earth method. When retained heights increase or restricting soil movements becomes important, propped embedded walls are used. In this case the free-earth method is adopted for design. For more detailed information see the Design guide for steel sheet piled bridge abutments(30). These two methods enable the embedment depth to be determined at the limiting equilibrium (failure) condition, and the anchor force where present. For design purposes a margin of safety is introduced.


70

15 September 2003


10.2 Methods of analysis for stability against overturning Various methods of analysis are in current use for assessing overall stability and determining depth of embedment for embedded retaining walls. They are all limiting equilibrium methods which are used for cantilever abutment walls, or abutment walls propped at or near the top. For design, the moment equation is directly or indirectly used to ensure that restoring moments exceed overturning moment by a prescribed safety margin. This is achieved by the introduction of partial factors or by a single factor of safety, termed a lumped factor. In the context of a limit equilibrium analysis, the factor of safety has two functions:

C

To make allowance for uncertainties in the evaluation of the soil parameter.

C

To ensure that deflections in service are not excessive.


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.

C

Limit equilibrium methods are widely used, hence there is a need for a factor to limit deflections.

C

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).

It is therefore recommended to use the Factor on Strength method using partial factors, and the Burland and Potts method based on a lumped factor of safety as a check.


71

15 September 2003



Discuss me ...

10.2.1

Factor on Strength method

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 φ '  and φ m'= tan − 1   Fc  Fφ 

where cm' and Nm' are the mobilised values of the respective strength parameters c' and N'. To maintain overall consistency

δm δ cmw ' cw = and = φm ' φ ' 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 ive 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


72

15 September 2003


where, Mfp = Mfa = Mw =

Moment of factored ive earth pressure Moment of factored active earth pressure Moment of net active water pressure

When the forces are expressed as functions of embedment depth do, this equation reduces to a cubic expression in 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) *

Serviceability Limit State

CIRIA 104(66)

Worst credible design

Effective stress

Total stress

1.2

1.5

Permanent works in stiff clays

1.2

-

Temporary works in stiff clays

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

1

Ground Properties tan N'

c'

cu

A

1.1

1.3

1.2

B1

1.0

1.0

1.0

C

1.25

1.6

1.4

Partial factors are applied on actions only.

Case A Case B Case C

10.2.2

is the loss of static equilibrium (strength of structural material or ground insignificant) is failure of structure or element (governed by the strength of the structural material) is failure in the ground.

Burland and Potts method

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.


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 ive

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 ive earth pressure diagram is modified by deducting the difference between the gross and modified active pressures from the gross ive pressure at any level. The modified active and ive 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 = = = =

moment of modified ive earth pressure moment of modified active earth pressure moment of net active water pressure lumped factor of safety

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.


74

15 September 2003


Table 10.3 Burland & Potts method factors


Reference

Design condition

Fp

Burland, Potts and Walsh

Ultimate Limit State

CIRIA Report 104(66)

Worst credible design, temporary works in stiff clay

1.0

Worst credible design, permanent works in stiff clays

1.5


75

1.5 - 2.0

15 September 2003


11

ABUTMENT WALLS - SOILSTRUCTURE INTERACTION

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 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.


11.1 Soil-structure interaction approach The introduction of soil pressures into structural design can give rise to some confusion over whether soil pressures are loads (in the sense that they are applied to a structural element) or are part of a structural element resistance (in the sense that they respond to imposed loads or displacements). The distinction is likely to be important in the understanding and treatment of integral bridges, where strain and displacements are continually varying, unlike normal ‘static’ situations. The distinction is also of significance to Limit State Design, since partial safety factors will generally be different for ‘loading’ and for ‘resistance’. Design rigour is needed to eliminate this potential confusion.

11.2 Mobilisation of earth pressure and soil-structure interaction The importance of soil structure interaction in design can best be shown by illustrating the mobilisation of earth lateral pressures as the embedded wall moves (see Figure 11.1, Ref. Potts and Fourie). Three modes of wall deformation, namely horizontal translation, rotation about the wall top and rotation about the toe, provided an analogy to commonly occurring cases. For embedded cantilever walls the mode of displacement is essentially one of rotation about the toe; for an anchor or bridge deck beam located near the top of the wall the mode is rotation about the top. Rotation about the base requires significantly more displacement to obtain failure conditions than do the other modes of displacement. For high values of the at-rest coefficient, Ko; active and ive conditions are mobilised at similar displacements. However, for low Ko values, active conditions occur before ive conditions. Clearly, displacements necessary to mobilise active and ive conditions are dependent on the value of Ko as well as on the mode of deformation. The limiting values of Ka, active, and Kp, ive, are less sensitive to the mode of displacement than the Ko value.


76

15 September 2003


Ko

Ka

Kp

K o = 2.0 1

4.0

∆

∆ H/2

∆ 1

2

1 3

2

Earth pressure coefficient K H

K o = 0.5

3.0

3 2.0 Ka Active side

Kp ive 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


Figure 11.1 Development of earth pressure coefficients with increasing wall displacement (rough wall) Deformation of flexible embedded sheet pile walls arising from soil-structure interaction will result in earth pressure distributions which cannot be predicted by classical earth pressure theories (see Figure 11.2). However, the study for the rigid wall case does show clearly actual behaviour rather than the assumed simplistic behaviour and makes the engineer appreciate the redistribution of earth pressure that will be created by flexible steel wall structures.

Actual measured pressure

Pressure profile from classical theory

Figure 11.2 Actual horizontal earth pressure distribution for a flexible sheet pile abutment

11.3 Soil-structure interaction analysis methods The soil-structure interaction approach models the earth pressure distribution which acts on the design configuration of the wall. The simplest of these soil-structure interaction methods are the Winkler(75) spring models where the soil is modelled as a spring. These are followed with increasing complexity by methods where the


77

15 September 2003


soil is modelled by Boundary element, finite difference and finite element numerical approximations.

11.3.1

Winkler spring models (beam on elastic foundations)


The beam or slab on an elastic foundation approach has found application in numerical analysis of sheet piles. Power series, finite differences, distribution and discrete element methods are employed for the solution of the governing differential equations. In each case the elastic foundation is assumed to generate reactive pressure proportional to the structural displacement (Winkler’s hypothesis). The soil response is usually characterised by a spring constant, which is related to the ‘coefficient of subgrade reaction’. Normally the coefficients of horizontal subgrade reaction recommended by Terzaghi(76) are used. The 'beam on elastic foundation' or 'subgrade reaction' approach is one commonly used for soil-structure interaction because of the ease with which it can be applied, and various methods are available. Commonly, the soil mass is modelled as a series of horizontal isolated springs, or as springs with some form of interconnection (see Figure 11.3). In addition, the in situ horizontal soil stresses are input together with the active and ive earth pressure coefficients to provide the limiting values of the horizontal effective stress. Where an anchor is to be simulated, additional springs at the required stiffness and at the appropriate level are input. Anchor

Slider

Spring

Figure 11.3 Spring model for analysis of abutment In most situations, the bending moments and shear forces obtained from the Winkler method are insensitive to the values of the spring stiffness chosen and used in the analysis. However, this is not the case for the prediction of deformations of the wall. Deformations obtained from these analyses can therefore only be regarded as rough estimates and need to be checked by field measurements. The analysis can be carried out with the wall being backfilled towards its top or excavation from the top of the wall downwards. If progressive softening, which takes place in the long term as a result of the swelling of clays, is to be modelled, then it is necessary to change the spring stiffnesses. Limitations of the ‘beam on elastic foundation’ approach are: C

Difficulty in determining appropriate spring stiffnesses for analysis.


78

15 September 2003


C

Method cannot directly simulate unusual initial soil stresses.

C

Development of wall friction and construction sequence.

C

Determining surface movements of the retained soil.

11.3.2

Boundary element and finite element methods

More sophisticated techniques are available which model general soil-structure interaction problems. These methods are more rigorous in the formulation of the problem and overcome the shortcomings of the beam on elastic foundations. A commonly used boundary element method is the one proposed by Pappin et al(77). Boundary Element methods still have shortcomings, therefore if a sophisticated analysis is required, a finite element analysis should be performed. These methods have the ability to predict both earth pressures and deformations with a minimum of simplifying assumptions. Where it is necessary to use finite element analysis techniques then specialist advice should be sought.


11.4 Global analysis of integral bridges In a non-integral bridge design the analysis of the bridge structure is performed by separate analyses of the deck structure and the substructure. For example, as in most cases the s are ‘pinned’, the bridge deck can be analysed using a grillage model and the reactions at the s then treated as loads acting on the substructure model. The substructure is analysed under the combined loading of the deck reactions and soil pressure loads. In an integral bridge, interaction between the deck and the substructure occurs with the adjacent soil becoming part of the structure’s resistance to deck loading. An effective design method that can be applied to integral bridges is one where the bridge deck and substructure models are separately analysed. Subsequently the models are modified to include the influence of the bridge deck to substructure connection and each model is re-analysed. The bridge deck to substructure connection is taken into by determining imposed displacements, rotations, forces and bending moments at the connection which results in a compatible set of imposed boundary conditions. These boundary conditions which are linear elastic translation and rotation springs, are applied both to the bridge deck and to the bridge substructure models.

11.5 Available soil-structure interaction analysis software There are a number of commercially available numerical analysis software products that are commonly in use by design practices. Three of the most well known are WALLAP, FREW and FLAC. Results obtained from these programs are very dependent on the relative stiffness of the chosen components of the model. The application of arbitrary rules for determining soil stiffnesses can result in large variations. Depth, soil type and over consolidation ratio can have considerable influence on the stress/strain properties of soils and these parameters should be obtained from proven experience. Programs


79

15 September 2003


need to be calibrated against measurements in well-documented case histories in retaining wall construction and worked examples made available to the .

11.5.1

WALLAP

WALLAP (Wall Analysis Program) is a widely used commercial package from Geosolve and described by Borin(78), designed specifically for routine retaining wall design. Finite elements are used to model the sheet pile wall and an elasto-plastic finite element model or a Winkler spring model can be chosen to represent the soil. The springs in the Winkler spring model can either be interconnected or independent. WALLAP can be used to model cantilever walls, anchored walls and strutted excavations. Initial or in situ soil pressures can be defined and the pressures determined during wall/soil movement are constrained to lie between active and ive limits. Wall excavation can be modelled together with de-watering, placing of surcharge and the introduction of struts and anchors. In addition complex water profiles can be included to model steady seepage, submergence and perched water tables.


The deliverables from the program include a stability analysis, wall displacement versus depth profile, bending moment, shear force, earth pressure distributions and strut loads. s 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

Winkler springs (subgrade reaction method).

C

Stiffness matrices obtained from pre-stored finite element analyses.

C

Mindlin method, where the soil is modelled as an elastic solid (constant stiffness with depth only).

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 s. 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

FLAC is a program for the solution of general geomechanical problems based on a finite difference method. The program has been developed by ITASCA Consulting Group, Minneapolis, USA. Amongst other uses, it is capable of solving a range of earth retention problems, and any type of non-linear soil stress/strain relationship can be followed. It is suitable for the solution of retaining wall, tunnel-lining and rockbolting problems.

11.6 Boundary conditions at the deck to abutment connection


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)

Figure 11.4 Soil-structure interaction diagram for WALLAP or FREW


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

θ y1 +

6 EI y L2

d z2 +

2 EI y L

θ y2

(1)

using the sign convention shown in Figure 11.5. θ y2

z

θ y1 my2

my1

y x

Figure 11.5 Sign convention For an excavation loadcase and for symmetric loading on the deck, rotation is symmetrical therefore: ˆ 2y2 = !2y1

and dZ1 = dZ2 = 0


which leads to an effective stiffness of the deck connection given by: ˆ

M y1

θ 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.

11.6.1 Modelling deck expansion Deck expansion using WALLAP can only be modelled by applying a preload to the prop at deck level. As prop preloads and stiffnesses are not proportional to member strain, i.e. g … P/AE if a preload is applied, the effective sequence of events modelled is: 1. 2. 3.

Prop stiffeners set to zero. Load applied. Prop stiffness reset to previous value.

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.


82

15 September 2003


Force Deck behaviour

Soil behaviour Force at equilibrium

Point of equilibrium

Displacement at equilibrium

Displacement

Figure 11.6 Equilibrium condition between deck and soil


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

Modelling braking loads

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.


83

15 September 2003


12

ABUTMENT WALLS - RESPONSE TO THERMAL DECK MOVEMENTS

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.

12.1 Bridge temperatures


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


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 s as they do so. In these cases simply ed bridges may actually be acting integrally. Table 12.1 Thermal strain design temperatures


Deck Structure type

Half range + 10oC

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.

12.2 Soil behaviour under cyclic loading It is widely recognised that soil behaviour under cyclic loading is controlled by: C

Soil type (cohesive or non-cohesive).

C

Magnitude of the shear strain, i.e. amplitude of induced movement or soil displacement.

C

Frequency and magnitude of loading.

C

Presence of ground water.

C

Initial density (for cohesionless soils).

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.


85

15 September 2003


Cyclic loading of soils due to thermal movement of a bridge superstructure occur at a much lower frequency (1 to 2 cycles per day for daily temperature changes and 1 to 2 cycles per year for seasonal changes). Under these conditions of comparatively slow cyclic loading the soil is likely to respond under drained or partially drained conditions with dissipation of pore pressures. It is quite probable that cohesive soils will behave in this manner. It is therefore anticipated that an increase in strength and stiffness of soil behind an integral abutment will take place if cyclic loading occurs.

12.3 Earth pressures due to wall displacement


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 underes 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 ive 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.


86

15 September 2003


Depth (m)

0

Wall

static cycle 1

2 K=2 4

e activ

dr hy

12 200

active

10

os

ta tic

8

tic ta os dr st h y at re

at re st

6

0

100

K=2 -100

-200

Lateral pressure acting on abutment (kPa)

Depth (m)

0

Wall

static cycle 1

2 4

12 200

active

0

100

tic ta os dr hy at rest

10

hy dr os ta tic

8

at re st

6

active


Figure 12.1 Lateral earth pressures - wall displacement towards fill

K=2 -100

-200

Lateral pressure acting on abutment (kPa)

Figure 12.2 Lateral earth pressures - wall displacement away from fill The results have shown that at serviceability limit state, the effect of cyclic ive 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 ive 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.


87

15 September 2003


12.4 Requirements of BA 42/96 In BA 42/96(9), the Highways Agency advises the limits on ive earth pressures and the value of Kp that it expects to be used in the calculation of ive earth pressure resistance for the cyclic loading caused behind bridge abutment walls due to thermal movement of the deck in an integral bridge. BA 42/96 states that the values of K* and Kp it gives are based on the findings obtained from experimental and analytical data, particularly that in the TRL Report by Springman et al(86), although this research applies specifically to short rigid walls and sand fill only. No adjustment is given for flexible steel walls, even though it is well known that earth pressure is considerably less, and of a redistributed form, behind such walls. Earth pressure coefficients are to be calculated according to BS 8002(36) but with the mobilisation and material factors as given in BA 42/96. The Eurocode 7(48) approach is also invoked.


Specific advice contained in BA 42/96 relating to soil behaviour is summarised below: 3.2 An increase of stiffness of granular soil occurs due to densification of the fill under the thermal cyclic movements induced by deck expansion. Even if the fill is placed in loose condition, representative c'peak and N'peak ... should be used throughout the design. 3.3 ... the ive earth pressure mobilised by a granular backfill on a [retaining wall] abutment of an integral bridge moving towards the backfill would act in an unfavourable manner. For this reason, the approach of Eurocode 7 ... is adopted in which the factor of M is 1/1.2, i.e a value <1, [and] is applied ... for earth pressure calculations. (BS 8002(36) gives a value of mobilisation factor, M = 1.2 to calculate active and at-rest earth pressure coefficients.) 3.3 ... Wall friction should be taken as * = design Nr/2 3.4 During displacement towards the backfill, ... Kp ... increases very rapidly at high angles of friction as follows: (see Table 12.2) Table 12.2 Values of Kp from BA 42/96(9) Kp Inclination of abutment backface N'

Vertical

20E forwards

20Ebackwards

30E

5

3

7

35E

6

4

12

40E

9

5

20

45E

15

6

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.


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

K ( ' (d/0.05 H)0.4Kp ... K* should not be taken as less than the ‘at rest’ earth pressure K0, nor less than Kp/3

3.5.4

3.6 Live load surcharge on backfill should be ignored when calculating the ive 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

0

2

4

6

8

10

12

14

16

Displacement, mm

Figure 12.3 Mobilised ive soil pressure coefficient, according to BA 42/96 Soil-structure interaction Unfortunately, BA 42 does not discuss the effects of soil-structure interaction, i.e. making allowance for the elastic strain in the deck when it exerts the force needed to displace the abutment into the soil. Indeed, the term “thermal displacement at the top of the abutment (d)” quoted in 3.5.3 of BA 42/96 is not defined or discussed. Unfortunately, a problem arises in trying to use BA 42 to determine the force to displace the wall from its at-rest position. The lower limit of a Kp suggests that there is no additional force needed to displace the wall until the displacement is sufficient to give a value of K*/Kp in excess of 0.33, which never happens for a typical abutment of bridges up to 60 m length. (The design displacements at the ends of a 60 m long bridge are 15 mm according to BA 42 - see Section 12.1).


89

15 September 2003


However, if the lower limit of Clause 3.5.4 is ignored, an interaction relationship of the type discussed in Section 11.6.1 can be considered. To illustrate this soil-structure interaction in an integral bridge, displacement/force relationships were calculated for a 60 m long composite deck with a 7 m high abutment wall, and this is illustrated in Figure 12.4. 8.0 7.0

Force (MN)

6.0

Response of deck (thermal strain 0.005)

5.0 4.0 3.0 2.0 1.0 0.0

Response of wall 0

5

10

15

Displacement (mm)


Figure 12.4 Response of deck and retaining wall in a typical bridge The force to displace the top of the wall is calculated on the basis of the K* value, the pressure diagram in Figure 3.2 of BA 42, and the assumption that there is no moment in the wall at the bottom of the retained height. The (FL parameter is omitted from both the force to displace the wall and from the force to restrain the wall against expansion. The soil is assumed to be well graded gravel; N is taken as 45E, a value which reflects compaction due to cyclic movement of the wall; the soil density is taken as 20 kN/m2; the deck area is taken to be 0.07 m2/m width. From Figure 12.4 it can be seen that there is very little restraint to thermal movement. The ‘equilibrium position corresponds to a movement of about 90% of the free movement (although the free movement is half what it would be for a nonintegral bridge fixed at the far end). The force in the deck is about 10% of the fully restrained force.

12.5 Comparison of Kp values in BA 42/96 with BS 8002 To clarify the design recommendations given in BA 42/96 on earth pressure, a comparison exercise was performed with values given by BS 8002. In BS 8002 the values of ive earth pressure coefficient Kp are obtained from the Rankine equation. For N'peak, Kp is given in Table 12.3.


90

15 September 2003


Table 12.3 Values of Kp and Ko given by BS 8002 based on N’peak N'peak

Kp

Ko

30

3

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.


Table 12.4 Values of Kp and Ko given by BS 8002, based on N’design N'

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:

M = 1.2, as Clause 3.2.5 of BS 8002, and thus N'design = N'/1.2

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:

Assume that N' is N'peak as per BA 42/96 Clause 3.2


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

5

42.0

34.7

35

6

45.6

40

40

9

53.1

45.2

45

15

61.0

50.2

N' obtained from Rankine and M from definition in BS 8002(36) Assume that N' is N'peak as per BA 42/96 Clause 3.2

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.


92

15 September 2003


13

STEEL PILES - AXIAL LOAD RESISTANCE

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.


13.1 Ultimate axial capacity and load transfer A piled bridge abutment or pier subjected to a vertical load, parallel to its longitudinal axis, will that load partly by shear generated over its length, due to the soil-pile wall friction or adhesion, and partly by normal stresses generated at the base or tip of the pile, due to end bearing resistance of the soil (see Figure 13.1.) P

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


93

15 September 2003


where: unit wall friction value 1) surface area of the pile in 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).

13.2 Vertical settlement and serviceability The design ultimate capacity of a steel pile, Rcd, is given by: Rcd =

where

Rsk

ξγ s Rsk Rbk (s (b .

+

= = = = =

Rsk

ξγ b ultimate wall resistance ultimate bearing resistance factor for wall friction resistance factor for base resistance factor to take into 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:


94

15 September 2003


(s (b .

= = =

1.3 1.3 1.5

Provided that

PDesign

≥ 1

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 Ultimate capacity in cohesive soils


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 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 all loads during construction. For these reasons it is common to calculate the base capacity of piles in clay in 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

cu is the undrained shear strength.


95

15 September 2003


As an indication, the undrained shear strength of cohesive soils is generally in the range between 20 kN/m2 for soft clay to 400 kN/m2 for very stiff or hard clays. For this range of cu, the end bearing values are in the range 0.2 to 3.6 MPa.

13.4 Ultimate capacity in cohesionless soils 13.4.1

Wall friction

The unit skin friction qs for a granular soil is given by: qs

=

where

2N N

13.4.2

=

Standard Penetration Test value

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.

Extensive experience with pile predictions based on the cone penetrometer in Holland has produced a set of design rules which have been summarised by Meigh(88). The magnitude of the end bearing capacity qb, generated in cohesionless soils can also be given by the relationship: qb

=

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.

13.5 Ultimate capacity in rock Where piles are driven through clay/sand strata but are terminated at depth into a relatively incompressible rock stratum, the main component providing resistance is the end bearing. In these cases low axial movements of the pile will occur, owing to the compressive strength of the rock, and it may not be possible to generate appreciable wall friction resistance in the clay/sand layers. In many cases, the maximum design load for such a pile will be determined by the stresses in the pile material itself, rather than the permitted end bearing capacity on the rock. P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

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 ted rock, end bearing values reduce significantly and can be of the order 10 to 100 MPa.

13.6 Mobilisation of wall friction on a retaining wall For a bridge abutment, adequate resistance must be provided by the steel pile to accommodate the vertical loads from the bridge deck superstructure. The vertical loads are applied as axial loads acting at the top of the pile. W

Active soil zone moving downwards


ive soil zone moving upwards

Figure 13.2 Generation of wall-soil friction by pile movement For design to resist the axial load acting at the top of the sheet pile, it is important that the overall behaviour of the pile abutment is considered. Although the design of the abutment to resist axial load is undertaken independently from the lateral loading case, the behaviour of the soil adjacent to the wall needs to be considered as the wall displaces laterally. The soil on the active or retained side of the wall moves down relative to the wall in order to mobilise friction in the beneficial direction and, on the ive side, the displaced soil has to move upward (see Figure 13.2). If the abutment itself displaces in a downward direction under the action of an axial load at the pile head, the wall friction on the active side will diminish. For an axially loaded pile, it may conservatively be assumed that wall friction resistance is mobilised along the wall bounded between excavation level and the pile tip (see Figure 13.3). Only the side of the wall in with the ive soil zone is then considered.


97

15 September 2003


W

Excavation level

Assumed length of wall providing wall friction resistance

Pile tip

Figure 13.3 Length of sheet pile contributing to wall friction

13.7 Determination of friction surface area


13.7.1

Retaining wall abutment

The surface area of sheet piles and High Modulus Piles can be obtained from British Steel’s 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

Tubular and box piles

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.


98

15 September 2003


13.8 Determination of end bearing area 13.8.1

Retaining wall abutment

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

Tubular and box piles

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 Buckling aspects of fully and partially embedded piles The wall thickness of the pile should be chosen such that diameter to thickness ratio D/t is sufficiently small to preclude local buckling at stresses up to the yield strength of the pile material. Guidance relating to local buckling can be found in API RP2A-LRFD(22). There are analytical solutions available to determine the buckling behaviour of fully and partially embedded piles but the methods are quite complex. One method is provided by Bowles(89). Bowles adopts the method of Wang(90) where the method is automated using a suitable computer analysis program. Although other methods are available (Davison and Robinson(91) and Reddy and Valsangkar(92)), the method proposed by Bowles is much easier to use.

13.9.1

Serviceability limit state

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.


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 γ fl γ f3 = where

P (fL (f3 Rcd ( .

Rcd

γξ

= = = = = =

unfactored axial load at the pile head partial factor for loads from BD 37(51) partial factor for innacurate assessment from BD 37(51) ultimate axial capacity of the soil resistance partial factors (s and (b factor to take into uncertainty of soil parameters determined on site or in the laboratory. R sk+ R


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)

Figure 13.4 Resistance versus pile head displacement The partial load factors are taken from BD 37(51) and the partial resistance factors from the draft version of Eurocode 7, (Table 7.1 and 7.2)(48). The equation above can be rewritten by inserting the magnitudes of the partial factors, therefore:

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.


100

15 September 2003


14

STEEL COLUMN-PILES - LATERAL LOAD RESISTANCE

Greimann and Wolde-Tinsae(93) provide a simplified design method for analysing piles in integral bridge abutments based on previous analytical models and observations of pile behaviour. Two failure modes for a pile were considered; the slip mechanism, where the pile slips through the soil; and the lateral mechanism, where the failure of the soil pile system is associated with lateral movement of the pile. Results predicted by the simplified model were compared to results from a non-linear finite element program and shown to be conservative. Where horizontal loading on individual piles is significant, pile analyses are required to determine the design bending moments. In certain situations, the interaction of individual piles will need to be considered and pile group analysis performed to predict the group load resistance and lateral movement.


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 in the design of piles.

14.1 Lateral loads from soil When an embankment is terminated adjacent to a piled bridge abutment, any underlying sub-soil stratum undergoing deformation may tend to ‘flow’ away from the embankment. Where this effect occurs, this effect results in lateral pressures which act on the pile shafts in addition to the other loads carried by the foundation. These types of piles are referred to as ‘spill-through’ piles. Where there is a likelihood of any soil-induced lateral loading being developed during or after the construction of a pile, for example, in river valleys with overlying soft clays, silts, etc. efforts should be made firstly to identify the extent of the problem and to determine the most economical solution. This could be either to take measures at the construction stage to reduce the magnitude of the effects or to design the piles to withstand the predicted lateral loading. Whatever option or combination of options are chosen, it is important that an appropriately detailed ground investigation is undertaken (see Section 9). For integral bridges, spill-through abutments reduce the soil induced lateral pressures, however, it is more difficult to obtain good compaction of fill material for backfilled abutments. Another option is to consider piled or un-piled bankseats. Where spill-through piled abutments need to be designed for soil-induced lateral loading, the procedure recommended in BA 25 Section 4.6 can be used.

14.2 Lateral forces at pile head design of piles for lateral loading is comprehensively covered in the CIRIA Report Design of laterally loaded bearing piles(67), publications by Poulos and Davies(94), Tomlinson(95) and the Steel bearing piles guide(12). These documents review the available methods for the analysis of laterally-loaded single piles and pile groups.


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 asg 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

Methods of analysis for ultimate or failure conditions

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

Serviceability and deformation analysis of piles

The performance of a laterally loaded pile in service can be predicted with reasonable confidence by using idealisations of soil stiffness. These include soil stiffness constant with depth, linearly varying with depth or by soil resistance-deformation curves (P-Y curves). Elastic continuum analysis Finite Element programs can also be used. These methods are fully explained in a CIRIA Report(67). Although numerous idealisation of soil stiffness analysis methods are available, the lateral deformation of the pile is best modelled using the Winkler Medium approach. This approach considers the pile to be a linear elastic beam ed by a series of discrete springs. A P-Y spring Winkler model for horizontal movement is shown in Figure 14.1. V M H

Figure 14.1 P-Y spring model for lateral pile resistance

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


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

Capabilities of available methods of analysis

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


Structural frame Unrealistic model (the soil is ignored).

Application

Output

End-bearing pile groups, with a small lateral load component (say up to 10% of the vertical load).

Axial load on piles is the only reasonable output.

Any laterally-loaded single pile or widely-spaced pile (s/D>3) group. The analysis can provide reasonable predictions for cyclic loading or for the development of plastic zones if suitable P-Y data are selected.

Depth, slope, moment and shear of the pile at any depth.

Winkler Medium or P-Y analysis

A reasonable model for single piles. However, inappropriate for pile groups with s/D < 8, because the continuity of the soil is not modelled.

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.

14.3 Analysis of pile groups Once a pile layout has been established for the bridge substructure and the initial estimate of the number of piles required has been determined using the predicted single pile design capacity from Section 13.2, the possible pile group effects will need to be considered. Piles in a group may be subject to the following effects: C

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.


103

15 September 2003


C

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.

Consideration of the ultimate failure mechanism of the foundation and incorporation of an overall reserve of strength for safety.

2.

Computation of the lateral translation and rotation of the foundation at working loads, and consideration of the effect of this deformation on the whole structure.

3.

Bending resistance of the piles.

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:


104

15 September 2003


Outline Pile Group Based on engineering assessment of - soils data - structural loading - construction constraints - possible pile types

Assess - reliability of data - sensitivity of structure

Select Method of Detailed Analysis Consider - available SI data - magnitude of lateral loads - complexity of pie data - batter of pipes

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

Appraise Performance of Foundation - servicability of structure - consider limit state criteria - global effects

Modify design as neccessary

Final design details

Figure 14.2 Analysis procedure for pile group


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.

Spring idealisation - In this method, the soil is modelled by an infinite number of discrete springs (Winkler Medium). Transfer of shear stresses within the soil mass is not modelled. The method has been extended (by Matlock and Reese(102)) to include non-linear springs, and it is generally referred to as the PY’ or the subgrade reaction method. The model is well developed, giving satisfactory predictions of the behaviour of single piles and pile groups. It is felt that the reservations expressed in CIRIA Report 103(67) in regard to pile groups is no longer valid because P-Y modifyers or ‘interaction factors’ which modify the relationship between load and displacement for each pile are now available. An example of the use of springs is described by Horsnell(103) where P-Y curves are used to represent the horizontal springs. Poulos describes the use of the elastic or elastic-plastic continuum approach which is incorporated in the program DEFPIG(104).

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


106

15 September 2003


programs are already available commercially, and it is anticipated that other comprehensive programs will be produced, to cater for most common practical problems. Two programs recommended in BA 25(?) include PGROUP (see Bannerjee et al(112)) and MPILE(113). The idealisation of an elastic continuum allows calculations to be performed which provide an insight into the behaviour of the pile group, and the sensitivity of the group to changing loading conditions and soil parameters. The application of these concepts is summarised in Table 14.1, together with the output of each method. The designer should select the method appropriate to the problem in hand, bearing in mind the complexity of the problem and the resources available. For the design of large piled foundations, an analysis based on the elastic continuum approach is considered to be one of the most satisfactory methods available at present, provide that the limit on displacement is satisfied.


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.

14.4 Behaviour of a spill-through column-pile abutment Column-pile abutments are recommended as one form of substructure which provides an attractive compliant to an integral bridge. Typically, a column-pile abutment consists of a single line of equally spaced tubular piles which are capped by a crosshead beam with an end screen wall. The lack of information for general abutment design (not specific to integral bridges only) prompted the Transport Research Laboratory to initiate long term monitoring of a spill-through bridge abutment with the objective in obtaining actual data relating to lateral soil pressures, response of the piles and the internal forces that were induced into the structure. This data provides an insight into the behaviour of a steel column-pile abutment configuration (see Section 5) as it illustrates that horizontal friction loads were generated in the PTFE sliding bearings. Columns for two spill-through abutments were monitored in collaboration with the University of Surrey for a bridge that carries the A3 trunkroad over the M25 motorway at Wisley(114). Each abutment is formed from six reinforced concrete columns spaced at 4 m centres ed on a continuous base slab. A transverse reinforced concrete capping beam provides the for the bridge deck beams and an endscreen wall is used to provide lateral to the soil immediately adjacent to it. A vertical cross-section through a column is shown in Figure 14.3. The design of the abutments for longitudinal bending was based upon an active soil pressure on the rear face of the capping beam and the columns, using a pressure coefficient Ka= 0.27. However, the lateral pressure on the columns was doubled to allow for the possible effect of side friction and arching of the soil since it was assumed the fill would push forward between the columns. Lateral pressures on the front face were ignored and the columns were assumed to bend as vertical cantilevers from the top of the base slab.


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)


Figure 14.3 Vertical cross section through column, Wisley bridge

End screen wall

8

Transverse capping beam

6 Resultant 4

Design

2

0

Pile

60

40

20

0

0 20 Lateral pressure (kPa)

Figure 14.4 Pressure profile throughout depth of abutment, Wisley bridge


108

15 September 2003


14.5 Integral bridges and crash resistance Since the advent of the Highways Agency’s standard BD 60/94, effective crash resistance design has become a critical aspect for intermediate pier and abutment s to bridges. Reinforced concrete columns have been known to shatter under vehicle impact (see photographs contained in BD 60/94). This is due to the dynamic effects of the impact. No design advice on these effects is given in BD 60/94, which reduces the accident loading to a pair of static forces. It is considered that the energy absorption aspect of vehicle impact is an important factor when considering the survivability of a and thereby of the overbridge itself, but this is not currently covered in the Highways Agency advice and Standards.


Since the mass of the bridge 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 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 element. A steel tubular column-pile is also very easy to install by driving and is thus a practical solution for bridge s. 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).


109

15 September 2003


15

COMPOSITE DECK DESIGN

The design of continuous composite decks for non-integral bridges is well understood and guidance is available(65). Such design includes: C

Consideration of the global analysis (usually by use of grillage analysis).

C

The design of the main girders to carry moments and shears.

C

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.

C

Selection of bracing to stabilise compression flanges at midspan (during construction) and adjacent to intermediate s (in service).

C

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.

15.1 Axial Loading It can be seen from the discussion in Section 12 that the thermal strain experienced by a bridge deck at ULS (a value of 0.0005 according to BA 42(9)) is very little affected by the typical abutment wall, even when the retained soil is well compacted granular fill. Consequently the axial stresses in the deck as a result of the partial restraint of expansion and contraction are small (of the order of 10 N/mm2 for the example quoted in Section 12.4). Since temperature loading is a combination 3 loading, it is more likely that a combination 1 loading will govern the deck design. If the soil is stiff, a larger axial force in the deck may arise due to HA loading. This is partly due to the restraint of bending of the wall by the soil, and partly due to the HA surcharge on the soil behind the abutment. However, if the deck were to act as a prop between two retaining walls, higher axial loads would result. The extreme case of this situation is when the deck is constructed before there is any excavation in front of the retaining wall (or alternatively, before any backfill is placed behind it). The wall then behaves as a propped cantilever rather than a free cantilever. In the SCI publication Integral steel bridges: Design of a single-span bridge - Worked example(2) it is shown that a total axial stress of over 60 N/mm2 results from the combination of forces due to propping of the wall and restraint of wall bending when HA load is applied. The only significant axial forces in the deck will thus be compressive. In midspan regions this will reduce the bottom flange stresses and increase the force in the slab. As the slab is rarely designed on the basis of axial strength, this additional axial stress should have little affect. However, at intermediate s, and at the end P:\CMP\Cmp657\pubs\P163\P163-Final.wpd

110

15 September 2003


s 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 ing 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.)

15.2 Moments due to frame action Where there is moment continuity between deck and wall (or column-pile), moments will be developed due to live loading and thermal displacements. If propped construction is used, or if moment continuity is created before all the deck load is added, there will be moments due to dead load as well. The governing design moments at the ends will be hogging moments, and this will cause compressive stresses in the bottom flange. This region will then have to be treated like the hogging moment regions adjacent to intermediate s, and bracing will be required to stabilise the bottom flange against buckling. Note that the axial loads discussed above will also contribute to the level of stress in the bottom flange and to the need for bracing. Where the beams are connected to a reinforced concrete capping beam, the moments transmitted between deck and wall will be transferred via shear connectors attached to each flange. There is a good case for making use of hooped connectors instead of the usual shear studs because, although they are more expensive to add during fabrication, they can transmit greater forces. Where the only significant moments are those due to live load, the fatigue design of the shear connection must be considered carefully; the moment due to the fatigue vehicle may be sufficiently large that fatigue resistance governs.


111

15 September 2003



16

REFERENCES

1.

THE STEEL CONSTRUCTION INSTITUTE Integral Bridges: Interim Report, Phase 1 Document RT 408, July 1994 Integral Bridges: Interim Report, Phase 2 Document RT 438, November 1995 (unpublished reports to British Steel)

2.

WAY, J.A. and YANDZIO, E. Steel integral bridges: Design of a single-span bridge - Worked example The Steel Construction Institute, 1997

3.

Steel integral bridges: Design of a multi-span bridge - Worked example The Steel Construction Institute (to be published)

4.

BURKE, M.P., Jr. Transportation Research Record, 1113 pp. 54-65 Bridge approach pavements, integral bridges and cycle-control ts Transportation Research Board, National Research Council, Washington D.C., 1987 and National Cooperative Highway Research Program, Synthesis of highway practice No. 141, Bridge deck ts, September 1989

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.

MOULTON, L.K., GANGARAO, H.V.S., and HALVORSEN, G.T. Tolerable movement criteria for highway bridges, pp. 1-109 Report FHWA/RD- 85/107 FHWA Research Development Technology, McLean, Va, 1985

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


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.

BURKE, M. P., Jr. tless Bridge Experience in the United States 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

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.

Correspondence between British Steel Sections, Plates & Commercial Steels and Edward P. Wasserman, Department of Transportation, State of Tennessee, Nashville, Tennessee, 1993

14.

The American Association of State Highway and Transportation Officials (AASHTO) Standard Specifications for Highway Bridges, 16th Edition AASHTO, Washington D.C., 1996

15.

HAMBLY, E. C. Bridge design for durability Integral bridge abutment details in practice and theory TRL, 1992

16.

WASSERMAN, E.P. tless bridge decks AISC Engineering Journal, Vol. 24, No. 3, 93-100, 3rd Quarter, 1987

17.

GREIMANN, L.F., WOLDE-TINSAAE, A.M., and YANG, P.S. Transportation Research Record 903: Skewed bridges with integral abutments Transportation Research Board, National Research Council, Washington D.C., 1983

18.

ROEDER, C.W. and MOORTY, S. Transportation Research Record 1290: Thermal movements in bridges Transportation Research Board National Research Council, Washington D.C., 1991


113

15 September 2003



Discuss me ...

19.

MOORTY, S. and ROEDER, C.W. Temperature-dependent bridge movements ASCE Journal of Structural Engineering, Vol. 118, No. 4, April 1992

20.

LOW, A. Integral Bridges - Designer’s 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 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.

AMERICAN PETROLEUM INSTITUTE API Recommended Practice 2A-LRFD (RP2A-LRFD) 1st Edition Recommended practice for planning, deg and constructing fixed offshore platforms - Load and resistance factor design API, 1993

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.

BRITISH STEEL SECTIONS, PLATES & COMMERCIAL STEELS Piling Handbook, Seventh Edition BS S, P & CS, 1997

25.

EUROPEAN COMMITTEE FOR STANDARDISATION Draft prENV 1993-5 Eurocode 3: Design of steel structures Part 5: Piling CEN, 1996

26.

INSTITUTION OF CIVIL ENGINEERS Specification for piling and embedded retaining walls Thomas Telford, 1996

27.

FEDERATION OF PILING SPECIALISTS Specification for steel sheet piling FPS, 1991

28.

EUROPEAN COMMITTEE FOR STANDARDISATION (CEN) Draft prEN 12699 Execution of special geotechnical works - displacement piles BSI, 1997


114

15 September 2003



Discuss me ...

29.

TECHNICAL EUROPEAN SHEET PILING ASSOCIATION Installation of steel sheet piles TESPA, 1993

30.

YANDZIO, E. Design guide for steel sheet piled bridge abutments The Steel Construction Institute, 1998

31.

United Kingdom Control of Pollution Act 1974 Section 40 HMSO, 1974

32.

BRITISH STANDARDS INSTITUTION BS 5228: Noise control on construction and open sites BSI, 1992

33.

CORNFIELD, G.M. Steel bearing piles, 4th Edition The Steel Construction Institute, 1989

34.

Proceedings of the International Conferences on the application of stress wave theory on piles Stockholm 1980, 1984; Ottawa, 1988; The Hague, Holland, 1992; Orlando, Florida, 1996 A.A. Balkema publishers, Rotterdam

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.

BRITISH STANDARDS INSTITUTION BS 8002: 1994 Code of practice for earth retaining structures BSI, 1994

37.

PADFIELD, C.J. and MAIR, R.J. Report 104: Design of retaining walls in stiff clays Construction Industry Research and Information Association (CIRIA), 1984

38.

BRITISH STANDARDS INSTITUTION BS EN 10248: Hot rolled sheet piling of non alloy steels BSI, 1996

39.

BRITISH STEEL SECTIONS, PLATES & COMMERCIAL STEELS The corrosion and protection of steel piling in temperate climates Publication P115, February 1994

40.

BRITISH STEEL SECTIONS, PLATES & COMMERCIAL STEELS The prevention of corrosion on structural steelwork Brochure reference SPCS 501 3 5/96, May 1996


115

15 September 2003



Discuss me ...

41.

BRITISH STANDARDS INSTITUTION BS 1377: 1990 Methods of tests for soils for civil engineering purposes BSI, 1990

42.

BRITISH STANDARDS INSTITUTION BS 5400: Steel, concrete and composite bridges Part 1: 1988: General statement Part 2: 1978: Specification for loads Part 3: 1982: Code of practice for design of steel bridges Part 4: 1990: Code of practice for design of concrete bridges Part 5: 1979: Code of practice for design of composite bridges Part 10: 1980: Code of practice for fatigue BSI

43.

BRITISH STANDARDS INSTITUTION BS 8004 :1984: Code of practice for foundations BSI, 1984

44.

BRITISH STANDARDS INSTITUTION BS 449: Specification for the use of structural steel in building Part 2: 1969: Metric units BSI, 1969

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.

POTTS, D. M. and FOURIE, A. B. A numerical study of the effects of wall deformation on earth pressures pp. 383-404 Int. Jnl. Num. and Anal. Meths in Geomech. Vol. 10, No. 3, 1986

47.

INSTITUTION OF STRUCTURAL ENGINEERS Soil structure interaction - the real behaviour of structures ISE, 1989

48.

BRITISH STANDARDS INSTITUTION Draft for Development DD ENV 1997-1: 1995 Eurocode 7: Geotechnical design Part 1: General rules (Includes the United Kingdom National Application Document) BSI, 1995

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


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.

INTERNATIONAL ORGANISATION FOR STANDARDISATION (ISO) 13819-2: Petroleum and Natural Gas Industries: Offshore structures Part 2: Fixed Steel Structures ISO, 1995

53.

BRITISH STANDARDS INSTITUTION BS EN 10 025: Hot rolled products of non-alloy structural steels Technical delivery conditions BSI, 1993

54.

AMERICAN PETROLEUM INSTITUTE API Specification 5L Specification for line pipe, 41st Edition API, 1995

55.

HEALTH AND SAFETY EXECUTIVE Offshore installations: Guidance on the design, construction and certification, Fourth Edition HSE, 1990

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.

CARD, G.B. and CARDER, D.R. Highways Agency/TRL Report 228 Movement trigger limits when applying the Observational Method to embedded retaining wall construction on highway schemes Transport Research Laboratory, 1996

59.

THE INSTITUTION OF CIVIL ENGINEERS The Observational Method in geotechnical engineering Thomas Telford, 1996

60.

CARDER, D.R., PRESS, D.J., MORLEY, C.H., and ALDERMAN, G.H. TRL Report 239: Behaviour during construction of a propped diaphragm wall founded in London clay at Aldershot Road Under Transport Research Laboratory, 1997


117

15 September 2003



Discuss me ...

61.

BRITISH STANDARDS INSTITUTION BS 7608: Code of practice for fatigue design and assessment of steel sections BSI, 1993

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.

HAMBLY, E.C. and OWENS, G.O. tless steel viaducts for railways Conference on Continuous and Integral Bridges E & FN Spon, 1994

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 ed composite bridges, 1991 The Steel Construction Institute

66.

PADFIELD, C.J. and MAIR, R.J. Report 104: Design of retaining walls in stiff clays Construction Industry Research and Information Association (CIRIA), 1984

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.

BRITISH STANDARDS INSTITUTION BS 5930: 1981 Code of practice for site investigations BSI, 1981

70.

HIGHWAYS AGENCY Design manual for roads and bridges, Volume 4, Section 1 HA 34/87: Ground investigation procedure The Stationery Office, 1987


118

15 September 2003



Discuss me ...

71.

WELTMAN, A.J. & HEAD, J.M. CIRIA Special Publication 25: Site investigation manual Construction Industry Research and Information Association (CIRIA), 1983

72.

UFF, J.F. & CLAYTON, C.R.I. CIRIA Special Publication 73: Role and responsibility in site investigation Construction Industry Research and Information Association (CIRIA), 1991

73.

MARTIN, W.S. CIRIA Special Publication 136: Site guide to foundation construction Construction Industry Research and Information Association (CIRIA), 1996

74.

BURLAND, J.B., POTTS, D.M. and WALSH, N.M. The overall stability of free and propped embedded earth retaining walls pp. 28-38, Ground Engineering,14(5), July 1981

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.

PAPPIN, J.W., SIMPSON, B., FELTON, P.J. and RAISON, C. Numerical analysis of flexible retaining walls Proceedings of the Numerical Methods in Engineering Theory and Application 85 Conference, Swansea, pp 789-802

78.

BORIN, D.L. WALLAP anchored and cantilevered retaining wall analysis program: ’s manual (Version 4) Geosolve (London), 1988

79.

BROOKS, N.J. and SPENCER, J.F. Design and recorded performance of a secant retaining wall in Croydon Proc. Int. Conf. Retaining Structures Cambridge, 1992

80.

PAPPIN, J.W., SIMPSON B., FELTON, P.J. and RAISON, C. Numerical analysis of flexible retaining walls Proc. Symp. Computer Applications in Geot. Engng. Midland Geot. Soc. Birmingham Univ. pp 195-212 1986

81.

PAPPIN, J.W., SIMPSON, B., FELTON, P.J. and RAISON, C. Numerical analysis of flexible retaining walls Symposium on Computer Applications in Geotechnical Engineering The Midland Geotechnical Society, UK, 1986


119

15 September 2003



Discuss me ...

82.

THE STEEL CONSTRUCTION INSTITUTE and BLACKWELL SCIENCE Steel designers’ manual - 5th edition Blackwell Science, 1994

83.

BROMS, B.B. and INGLESON, I. Earth pressure against the abutments of a rigid frame bridge pp. 15-28, Geotechnique 21, No.1, 1971

84.

BROMS, B.B. and INGLESON, I. Lateral earth pressure on a bridge abutment Vol.1, pp 117-123. Proc. 5th European Committee on Soil Mechanics and Foundation Engineering, Madrid, 1972

85.

BROMS, B.B. and INGLESON I. Written discussion - Lateral earth pressure on a bridge abutment; Lateral earth pressures on walls and measurements under different temperature conditions Reprint and Preliminary Reports, No. 52, Swedish Geotechnical Institute, Stockholm, 1973

86.

SPRINGMAN, S.M., NORRISH, A.R.M. and NG, C.W.W. TRL Report 146: Cyclic loading of sand behind integral bridge abutments Transport Research Laboratory, 1996

87.

OASYS Geotechnical suite of programs PILE: - Vertical pile capacity Ove Arup & Partners (London), 1991

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 pp. 275-294, JSD, ASCE, Vol. 93, ST 1, Feb. 1967

91.

DAVISON, M.T. and ROBINSON, K.E. Bending and buckling of partially embedded piles Vol. 2., pp. 243-246, Proc. 6th International Conference on Soil Mechanics and Foundation Engineering, 1965

92.

REDDY, A.S., and VALSANGKAR A.J. Buckling of fully and partially embedded piles pp. 1951-1965, J. Soil Mechs. and Founds Div., ASCE, Vol. 96 (SM6), November 1970


120

15 September 2003



Discuss me ...

93.

GREIMANN, L.F. and WOLDE-TINSAAE, A.M. Design model for piles in tless bridges Journal of Structural Engineering, Vol. 114, No. 15, 1987

94.

POULOS, H.G. and DAVIS, E.H. Pile foundation analysis and design John Wiley and Sons, 1980

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.

REESE, L.C. and COX, W.R. Offshore Technology Conference, Houston, Texas, USA OTC 2312 Field testing and analysis of laterally loaded piles in stiff clay May 1975

100.

O’NEILL, M.W. and MURCHINSON, J.M. An evaluation of P-Y relationships in sands A report to the American Petroleum Institute, May 1983

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.

MATLOCK, H. and REESE, L.C. Generalised solutions for laterally-loaded piles Proc. Am. Soc. Civ. Engrs. - J. Soil Mech. Found. Div., Oct. 1960, 86(SM5), 63 to 91

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


121

15 September 2003



Discuss me ...

104.

POULOS, H.G. ’s 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.

BANNERJEE, P.K. and DRISCOLL, R.M. Three-dimensional analysis of raked pile groups Proceedings of the Institution of Civil Engineers Dec. 1976, 61(2), 635 to 671

111.

BUTTERFIELD, R. and DOUGLAS, R.A. CIRIA Technical Note 108: Flexibility coefficients for the design of piles and pile groups Construction Industry Research and Information Association (CIRIA), 1981

112.

BANNERGEE, P.K., DRISCOLL, R.M and DAVIES, T. PGROUP - Program for the analysis of pile groups of any geometry subjected to horizontal and vertical loads and moments Manual HECB/B/7, Highways Computing Division of the Dept. of Transport, 1981

113.

RANDOLPH, M.F. and RANPURIA, A.K. MPILE: - Program for the analysis of pile groups under general three dimensional loading conditions Manual HCSL/B/35, Highways Computing Division of Dept. of Transport, 1984


122

15 September 2003



Discuss me ...

114.

LINDSELL, P. and S.H. BUCHNER, S.H. Long-term monitoring of spill-through bridge abutments IStructE/BRE Seminar on Structural Assessment - Based on full and large scale testing, Watford, April 1987

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.

AMDAHL and FRIEZE Conference on accidental loadings on marine structures: Risk and response Royal Institute of Naval Architects (RINA), London, 1996


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

Integral Bridges 4b386t

Overview 4q3b3c

More details 26j3b

Related Documents 171j1w

Integral Bridges 4b386t

Integral Bridges 4b386t

Integral Bridges 4b386t

Pca - Prestressed Beam Integral Bridges 41t53

Integral Abutments For Prestressed Beam Bridges 6c4i68

Integral Concrete Bridges To Eurocode 2 5u5go

More Documents from "James Mcguire" 5r215

Granor Elastomeric Bearings Pads Strips - Series Bs 3s3zt

Integral Bridges 4b386t

Unit Weight Of Materials Used At Construction Site 11546d

Gea 2h Vertical Flow Fills Plasdek Biodek Product Profile 3ql1c

Gb Gea 2h Product Profile Biodek Cross-fluted Fills 2r1j4e

Matriz Comparativa 7605