Continuous and Integral Bridges by B. Pritchard

Continuous
and Integral
Bridges
Other books on bridge s from Taylor & Francis
Bridge Bearings and Expansion Joints

D.J. Lee
Bridge Deck Behaviour
E.C. Hambly
Bridge Management
Inspection, maintenance, assessment and repair
Edited by J.E. Harding, G.A.R. Parke and M.J. Ryall
Concrete Bridge Designer's Manual
E. Pennells
Concrete Bridge Engineering: Performance and Advances
Edited by R.J. Cope
Creep and Shrinkage of Concrete
Edited by Z.P. Bazant and I. Carol
The Design Life of Structures
Edited by G. Somerville
Developments in Structural Engineering
Proceedings of the Forth Rail Bridge Centenary Conference
Edited by B.H.V. Topping
The Maintenance of Brick and Stone Masonry Structures
Edited by A.M. Sowden
Prestressed Steel Bridges
Theory and Design
M.S. Troitsky
Santiago Calatrava
Edited by D. Sharp
Tubular Structures V
Edited by M.G. Coutie and G. Davies
For full details, contact The Promotion Department, Taylor & Francis,
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN, Tel: 071-865 0066,
Fax: 071-522 9623.
Continuous
and Integral
Bridges
Proceedings o f the Henderson Colloquiu m
T o w a r d s Joint-Fre e Bridges '
organized b y th e Britis h Grou p
of the International Associatio n
for Bridge and Structura l Engineerin g
Pembroke College, Cambridge, UK

20-21 July 1993
Edited b y
BRIAN PRITCHARD
Consultant to W.S. Atkins Consultants Ltd
and Colebrand Ltd,
Haywards Heath, UK
Taylor & Francis

Taylor & Francis Group
LONDON AND NEW YORK
Published b y Taylo r & Francis
2 Par k Square, Milton Park , Abingdon, Oxon , 0 X 1 4 4RN
2 7 0 Madiso n Ave , Ne w Yor k NY 1001 6
First editio n 199 4

Reprinted 199 5
Transferred to Digital Printin g 2 0 0 6
© 199 4 Taylor & Francis
ISBN 0 4 1 9 1903 0 9
Apart fro m an y fai r dealing fo r th e purpose s o f researc h or private study , o r

criticism o r review , a s permitte d unde r th e U K Copyrigh t Design s an d Patent s
Act, 1988 , this publicatio n ma y no t b e reproduced , stored , o r transmitted, i n
any for m o r by an y means , withou t th e prio r permission i n writin g o f th e
publishers, o r i n th e cas e o f reprographi c reproduction onl y i n accordanc e with
the term s o f th e licence s issue d by th e Copyrigh t Licensin g Agenc y i n th e UK ,
or i n accordanc e with th e term s of licence s issue d b y th e appropriate
Reproduction Right s Organizatio n outsid e th e UK . Enquirie s concernin g
reproduction outsid e th e term s state d her e shoul d be sen t t o th e publisher s a t the
London addres s printe d o n thi s page .
The publishe r make s n o representation , expres s o r implied, wit h regard to th e
accuracy o f th e informatio n containe d i n thi s boo k an d canno t accep t any lega l
responsibility o r liability fo r an y error s o r omissions tha t may b e made .
A Catalogu e recor d for thi s boo k i s availabl e fro m th e Britis h Librar y
Library o f Congres s Cataloging-in-Publicatio n Dat a availabl e
Publisher's Not e
This boo k ha s bee n proudced fro m camer a ready cop y provide d b y th e
individual contributors .
The publisher ha s gone t o great lengths t o ensure th e quality o f this reprint
but points ou t that some imperfection s i n th e original ma y be apparen t
Contents
Contributors and participants i x

Preface xi i
Introduction xii i
B.P. PRITCHARD
PART ON E PROBLEM S O F BRIDGE ARTICULATION 1
1 Performanc e of deck expansion joint s 3

I.D. J O H N S ON
2 Researc h requirements for bridge deck expansion joint s 1 1

J.R. C U N I N G H A ME
3 Bearin g performance 2 1
J. L U ND
4 Effect s o f temperature , shrinkage an d cree p o n integra l

bridges 3 3
B.A. NICHOLSON
Report on Session One 4 1

A.M. L OW
PART TW O ADVANTAGE S O F BRIDGE CONTINUITY 4 3
5 Historica l revie w o f th e developmen t o f continuit y an d

expansion joint s in bridges 4 5
CP. TILLY
6 Advantage s o f bridge deck continuity 5 5

B.P. PRITCHARD
7 Continuit y i n decks with precast beams - practica l issues 6 5

H.P.J. T A Y L OR
Report on Session Tw o 72
S. S H A N M U G AM
vi Contents
PART THRE E CONTINUOU S BRIDGE S
8 Desig n concept of the Trans-Toky o Ba y Bridg e

Y. SHIOI
S. N A K A M U RA
9 Continuou s steel and composit e bridges

J.D. P A R S O NS
Report o n Sessio n Thre e

G.P. TILLY
PART FOU R INTEGRA L BRIDGE S
10 Danis h integral bridges

E.M. VEJE
11 Precas t segmental constructio n

J. M U L L ER
12 Italia n joint-free bridge decks

F. MARTINEZ Y CABRERA
13 Th e U K Departmen t o f Transpor t vie w o n continuity /

integral bridges
G.P. PATERSON
14 Jointles s steel viaducts for railwa y

E.C. H A M B LY
G.W. O W E NS
Report o n Sessio n Four

S. S H A N M U G AM
PART FIV E LONG-LENGT H CONTINUIT Y
15 Th e desig n an d performanc e o f continuousl y reinforce d

concrete pavement s
F. FUCHS
R. D U M O NT
16 Experienc e i n th e us e o f continuou s welde d rail s o n

The Netherland s railway network
G. BIER
Contents
17 Continuou s arche s
D.R. T H O M P S ON and G.W. CLARK
Report on Session Five

B.P. PRITCHARD
PART SI X RETROFITTIN G
18 Stud y on jointless bridg e construction in Japa n

H. T A K A NO
19 Semi-integra l bridges : a concept whose time has come?

M.P. BURKE, Jr
Report on Session Si x
D.W. QUINION
PART SEVE N PRECAS T BEA M DEC K CONTINUITY
2 0 Continuit y o f precas t pretensione d concret e beam s b y

embedment i n integral crosshead s
B.P. PRITCHARD
21 Locall y separate d dec k sla b continuit y i n composit e

bridges
A. K U M AR
Report on Session Seven

G.P. TILLY
PART EIGH T SOIL-STRUCTUR E INTERACTION
22 Soil-structur e interaction : centrifug e modellin g o f

integral bridge abutment s
S.M. SPRINGMAN and A.R.M. NORRISH
23 Th e contributio n o f th e roa d pavemen t i n absorbin g

horizontal movements i n an embankmen t
A.M. L OW
viii Contents
2 4 Th e interactio n betwee n bridge s an d embankment s
and the use o f run-on slabs 27 3
S: T H O R B U RN
Report on Session Eight 28 7

D.W. QUINION
Concluding summar y 28 9
A.M. L OW
Author index 29 1
Subject index 29 3
Contributors and
participants
Ir G. Bier
Ingenieursbureau, Nederlands e Spoorwegen , Utrecht , The Netherland s
Mr M.P. Burke J r
Burgess & Niple Ltd, Columbus, Ohio , US A
Mr G.W. Clark
British Rail Research, Derby, U K
Mr J.R . Cuningham e
Transport Research Laboratory, Crowthorne, UK
Dr J.W. Dougil l
Institution o f Structural Engineers, London , U K
Mr R. Dumon t
Ministry o f Infrastructure an d Transport , Wallonia, Belgiu m
Mr M. Fuch s
Belgian Roa d Research Centre, Brussels, Belgiu m
Mr A. A. Hajiarshe d
Department o f Civil Engineering , Universit y o f Nottingham, U K
DrE.C. Hambl y
Edmund Hambl y Ltd , Berkhamsted, U K
Mr I.D. Johnso n
G Maunsell & Partners, Birmingham , U K
Dr A. Kumar
Kumar Associates, Beaconsfield , U K
Mr A.M. Lo w
Ove Aru p & Partners, London , U K
Mr J . Lun d
Allott & Lomax, Manchester, U K
Professor F . Martinez y Cabrera
Department o f Structural Engineering, Polytechni c o f Milan, Milan , Ital y
Dr C. Middleto n
Department o f Engineering, Cambridg e University, Cambridge , UK
Mr R.J.W . Miln e
Hon Secretary , IABSE British Group, London, U K
x Con trib u tors an d participan ts
Mr J . Mulle r
Jean Muller International , Montigny l e Bretonneux , Franc e
Dr S. Nakamur a
Nippon Steel Corporation, Kobe, Japa n
DrC. N g
Department o f Engineering, Universit y o f Cambridge, Cambridge, UK
Mr B.A. Nicholso n
Consulting Engineer, Bicester , UK
Ms A.R.M. Norris h
Dr G.W. Owen s
Steel Construction Institute , Ascot , Berkshire , UK
Mr J.D. Parson s
Cass Hayward & Partners, Chepstow, U K
Mr G.P . Paterso n
Bridges Engineerin g Division , Department o f Transport, London, U K
Mr A.J . Powderha m
Mott MacDonald, Croydon , U K
Mr B.P. Pritchard
Consultant t o W.S. Atkins Consultants Lt d and Colebrand Ltd,
Hay wards Heath , UK
Mr D.W. Quinio n
Chairman, IABS E British Group, Salisbury, U K
Mr S . Shanmuga m
Bridges Engineerin g Division , Department o f Transport, London, U K
D r Y . Shio i
Trans-Tokyo Bay Highway Corporation, Japan
Dr S.M. Springma n
Mr H. Takan o
Engineering Department , Metropolita n Expresswa y Public Corporation ,
l o k y o , Japa n
Dr H.P.J. Taylo r
Costain Buildin g Products Ltd, Stamford, U K
Dr D.R. Thompso n
British Rail Research, Derby, U K
Dr S. Thorbur n
Thorburn Ltd, Glasgow, Scotland, U K
Contributors and participants
Dr G.P. Tilly
Gifford & Partners, Southampton , U K
Mr E.M . Vej e
COWIconsult, Lyngby , Denmar k
Preface
Welcome t o th e proceeding s o f th e Britis h Grou p o f IABSE' s annua l

colloquium whic h thi s yea r concerne d joint-fre e bridges . Th e firs t o f
these colloqui a wa s held in 1974 during the Chairmanship of Dr William
Henderson fro m 1967 to 1979. In appreciatio n o f hi s contributio n th e
British National Grou p has named the m 'Henderso n Colloquia \
The subjec t fo r a colloquiu m i s selecte d a s bein g o f curren t concern
and worthy of debate over 2V2 days by some 3 0 contributors cloistered i n
Pembroke College, Cambridge . The contributors are invited o n the basi s
of their experiences an d the views they are likely to have, and come fro m
several countrie s t o maximiz e th e benefit s t o b e gaine d b y all . Th e
informal arrangement s encourag e discussio n an d debat e an d als o
provide introduction s t o ne w colleagues . Th e subject s i n th e pas t hav e
dealt wit h desig n methods , materials , constructio n practices , profes -
sional liabilit y an d wit h type s o f structure , suc h a s surface-stresse d
building structures in 1992. Th e proceeding s ar e published t o make th e
information an d view s expresse d availabl e t o al l intereste d peopl e an d
are late r discussed a t a joint meetin g organize d b y th e Britis h Grou p o f
IABSE and the Institution o f Structural Engineers.
On behalf of the British Group of IABSE I thank all the participants for
their contribution s t o th e colloquiu m an d th e spiri t o f camaraderi e
which developed . W e ar e al l gratefu l fo r th e preparatio n o f th e
colloquium b y Bria n Pritchard an d hi s organizin g committe e a s well a s
Bob Milne wh o mad e i t all happen .
D.W. Quinio n
Chairman, British Group, IABSE
Introduction
The late Bernard Godfrey, fondly remembered as a very active member of

the (then ) British Constructional Steelwork Association , beside s IABSE,
visited America in 1989 . H e returned full of enthusiasm fo r the example s
of the new joint-fre e typ e bridge s h e ha d come across , particularly som e
very lon g viaduct s i n Tennessee . A t th e nex t IABS E Britis h Grou p
meeting h e propose d tha t the 199 3 Henderson colloquiu m shoul d be o n
the subjec t o f joint-fre e bridges . Bernard' s proposa l wa s immediatel y
accepted an d h e starte d t o recrui t his committee . Sadly , Bernar d die d
several month s later .
Responding t o Bernard's enthusiasm, a programme entitled Toward s
joint-free bridges ' wa s pu t togethe r b y a n organizin g committe e
consisting of :
B.P. Pritchar d (Consultant , W.S. Atkins Consultant s Ltd ) Chairman
A.M. Lo w (Ov e Arup & Partners)
G.A. Paterso n (Department of Transport)
D.W. Quinio n (Chairman , British Group, IABSE)
Dr G.P. Tilly (Giffor d & Partners Ltd)
R.J.W. Milne (Institutio n o f Structural Engineers) Secretary
The first five named individual s als o acted as chairmen and reporters for
various sessions . M r S . Shanmuga m (Departmen t o f Transport ) als o
helped by reporting on tw o o f the sessions .
The natur e o f th e colloquiu m i s unique , i n tha t attendanc e i s b y
invitation only , wit h each of the 3 0 delegates, includin g representative s
from eigh t oversea s countries , contributin g a paper . Onl y shor t
introductions ar e give n b y author s an d th e mai n purpos e o f th e
colloquium i s hopefull y fulfille d b y th e length y discussio n period s
allocated afte r each session .
Bernard Godfrey' s enthusias m reflecte d th e growin g world-wid e
interest in joint-free bridging. Joint s are a continuing proble m i n existin g
bridging no t jus t becaus e o f thei r ow n failure s an d maintenanc e
problems, bu t becaus e o f th e enormou s amoun t o f corrosio n damage ,
measured i n billion s o f pounds, whic h ca n be cause d t o the underlyin g
substructures b y leakag e throug h th e joint s o f dec k run-of f wate r
containing corrosiv e de-icing salts. It is not, therefore, surprising that the
UK Departmen t o f Transport ar e currently favouring multi-spa n bridg e
deck continuit y and , indeed, joint-fre e integra l bridges u p t o some 7 0 m
xiv Introduction
long. I t is als o eviden t tha t longer an d longe r integra l bridges ar e being

constructed i n Europe and America.
With thes e factor s i n mind , th e organizin g committe e invite d
representatives from all sectors of the bridging industry and particularly
from countrie s heavil y involve d i n integra l bridg e construction . Afte r
much deliberation, it was decide d t o divide the two-day colloquium int o
eight sessions , havin g th e following themes :
1 Problem s of bridge articulation
2 Advantage s of bridge continuit y
3 Continuou s bridge s
4 Integra l bridges
5 Long-lengt h continuit y
6 Retrofittin g
7 Precas t beam deck continuit y
8 Soil-structur e interaction
In th e event , th e colloquiu m attracte d excellen t paper s i n al l sessions ,
with livel y an d informe d discussio n inevitabl y exceedin g th e tim e
allotted.
It i s hope d tha t thi s boo k convey s thes e qualities , whic h wer e ver y
evident t o thos e wh o attende d th e colloquium . I t canno t convey ,
however, th e spiri t o f th e even t - a head y mixtur e o f internationa l
camaraderie an d exper t first-han d knowledg e overlai n wit h th e
nostalgia for student day s which Cambridg e generates.
May I conclude thi s introductio n by recordin g my heartfel t thanks t o
the participant s an d t o th e hard-workin g member s o f th e organizin g
committee, wh o al l contribute d so much to this successfu l event . I must
also thank Bernard, no doub t in a bridge-decorated afterlife, for pointin g
the way .
Brian Pritchard
P A R T ON E
PROBLEMS
OF BRIDG E
ARTICULATION
1 PERFORMANC E O F
DECK E X P A N S I O N JOINT S
I.D. J O H N S ON
G. Maunsell & Partners, Birmingham, UK
Abstract
The infrastructure developments of the 1960's led to a considerable
increase in the UK bridge stock. It is recognised that the maintenance
of this stock now represents a major annual expenditure which has been
increased by the combination of the use of rock salt for de-icing and
the deficient performance of expansion joints.
Due to concern expressed by the Department of Transport regarding
bridge deck expansion joints the Transport Research Laboratory
appointed G. Maunsell & Partners to carry out a study of the condition
and performance of 250 highway bridge deck joints in-service. The
commission was awarded in February 1990 and completed in July 1992.
This paper is based on the findings of the study and discusses the
selection, specification and detailing of joints and associated
components.
Inspections of the joints indicated that a substantial proportion
has some form of defect and that a significant proportion leak.
Bearing shelves and sub-surface drainage systems were found to be
poorly detailed or badly maintained in a large number of cases and this
exacerbated the effects of the leakage of run-off water through
expansion joints.
Keywords: Bridge, Expansion Joint, Performance, Selection,
Specification, Detailing.
1 Background
Several studies have been conducted over the last decade into the
durability of highway bridges, Booth et al (1987), Weisgerber et al
(1987), PIARC (1987) and Wallbank (1989). All of these studies,
regardless of the country involved, highlight expansion joints as being
critical elements affecting the maintenance-free lives of structures
and include recommendations that specific studies be conducted on the
performance of joints in-service. The most recent national
investigations of expansion joints in the UK prior to this study were
conducted by Price (1984).
Until the mid 1970's the most common forms of expansion joint in use
in the UK were the buried and nosing types, however, as a result of the
failure of the surfacing over buried joints and the generally poor
Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published i n 1994

by Taylo r & Francis, 2 Park Square, Milton Park , Abingdon, Oxon , OX1 4 4RN. ISBN: 0 419 1903 0 9.
4 Johnson
performance of nosing joints, the asphaltic plug joint (APJ) has now
been widely adopted.
The original APJs, developed in Italy and the UK during the late
1970's, suffered from an unacceptable degree of tracking during periods
of hot weather and, although there have been continuing modifications
of the bitumen binders by the manufacturers, it is noticeable that the
present day performance of APJs is not wholly satisfactory.
The work by Price included a limited survey of APJs. The
significant annual expenditure now associated with the maintenance of
the UK bridge stock, coupled with the increased use of APJs, and the
identification of the performance of expansion joints as critically
affecting the durability of highway bridges lead to the Transport
Research Laboratory appointing G. Maunsell & Partners to conduct a
study of the condition and performance of expansion joints in-service.
The study was intended to update the available knowledge concerning
the performance of the more commonly used types of joint, namely APJs,
reinforced elastomeric joints (REJs) and elastomeric in metal runner
(EMR) joints and involved the inspection of 250 joints over a period of
18 months. Full details of the joint types studied, procedures used
and the locations of the joints surveyed are included in the final
report, Johnson and McAndrew (1993).
It was found that the majority of expansion joints have some form of
defect and that the problems identified by previous workers still
persist. This paper is based on the findings of the study and
discusses the selection, specification and detailing of joints and
their associated components.
2 Selection
2.1 Performance requirements

The purpose of expansion joints is to permit movement of the bridge
deck whilst ensuring that the ride quality and watertightness of the
joint are not unacceptably affected. Furthermore, it is important that
the joint be able to accommodate freely the design movements expected
of the structure, whether they be rotational, lateral or longitudinal,
without either transferring potentially damaging forces into the
structure or causing distress to the joint itself.
As well as these performance requirements, the process of selection
of an appropriate joint should also take into account the
characteristics of the site that is being considered. Although the
Department of Transport (DoT) Departmental Standard and associated
Advice Note provide some guidance on the selection of joints, DoT
(1989a), DoT (1989b), it could be argued that the implementation of the
current UK Specification for Highway Works, DoT (1992), limits the
selection process and does not necessarily enable to engineer to
specify the most durable or appropriate joint type for a given
situation. DoT policy with respect to the assessment of durability is
currently under review and it is understood that the concept of whole
life costing is to be introduced, this point is discussed further on in
this paper.
Performance of deck expansion joints 5
2.2 Movement
Acceptable movement ranges for the various joint types in use on UK
roads are given in the Departmental Standard, however, the relative
proportions of static and dynamic components can significantly affect
the performance of a joint. The acceptability limits provided
for these components in relation to joint performance, particularly
with respect to APJs, is less clear and is open to some conjecture.
The surveys conducted as part of the study identified that the
performance of APJs was often incompatible with the relatively large
magnitude of dynamic, or high frequency, displacements and rotations
that can be experienced on steel composite decks when these structures
are subjected to heavy traffic densities, particularly those
incorporating a high proportion of heavy goods vehicles (HGVs). It was
noted, however, that the cracking and leakage of APJs resulting from
this incompatibility was usually confined to untrafficked areas, such
as the hardshoulder or central reserve, and, to a lesser extent, to the
outer lane of dual 3 lane carriageways. It is considered that this
distribution of defects is due to the remoulding effect that
trafficking, particularly by HGVs, can have on the binder matrix of
APJs.
It was reported by maintenance engineers that cracking of APJs was
often initiated after the occurrence of hard frosts. The onset of this
cracking is considered to be the result of relatively sudden thermal
movements associated with rapid temperature changes.
Cracking, due to either trafficking or frost, is believed to occur
due to limitations in the elasto-plastic behaviour of the binder
material. It is, however, considered that insufficient information is
currently available to enable practical guidelines to be prepared which
would then permit the design engineer to assess potential joint
performance from the binder properties.
Concrete decks generally tend to be stiffer than their steel
counterparts, the dynamic component of the total joint movement is,
therefore, not as onerous. It is, however, essential that the static,
or low frequency, movements arising from the creep and shrinkage of
concrete decks are fully assessed; inadequate provision for these
movements has been found to result in the dislocation of elastomeric
inserts in EMR joints. Similar errors could also cause distress in
REJs and may lead to splitting. Movements due to creep and shrinkage
appear to be significant only on relatively long span, or multi-span
continuous structures. No deleterious effects in APJs have been
attributed to these movements.
2.3 Influence of site on joint type

The performance requirements of a joint are partially related to the
characteristics of the site in which it is to be placed and the site
should, therefore, influence the selection procedure. Moreover, the
nature of the installation itself, whether new construction or
replacement work, can place significant restrictions on the suitability
of some joint types.
Heavy trafficking is a primary concern and can lead to tracking of
APJs and an unacceptable ride quality. If tracking is allowed to
develop unabated then increased impact loads can result in the
6 Johnson
deterioration of the adjacent surfacing and the eventual potholing of

joints. Because of this, it is recognised by many maintenance
authorities that regular resurfacing of the upper 40mm of APJs may be
required on busy routes.
The behaviour of REJs and EMR joints can also be adversely affected
by trafficking. The inspections carried out indicated that bolted
connections, and particularly those anchored by the placement of a
mortar bed under the complete joint, are inappropriate for routes with
a high flow of HGVs. The vibrations caused by such vehicle flows have
been found to result in the stripping of bolt threads and, in some
cases, severe deterioration of the anchorage material. It should be
noted that an anchorage bedding has been observed to reach the point of
disintegration within a period of three months from the time initial
signs of distress were noted; this highlights the need for the regular
inspection of joints to assess their integrity.
In addition to vibration related damage, mechanical anchorages are
also subjected to dynamic impact loads which have been shown to be up
to 1.7 times the static wheel load, Koster (1986). These impact loads
no doubt play a significant role in the deterioration of anchorage
beddings as noted above. Such deterioration is not as frequent in the
anchorages of EMR joints, the majority of which are cast-in and have a
reinforcement cage integral with that of the deck.
Cast-in anchorage details are, however, usually wholly inappropriate
for repair situations where a failed joint is to be replaced. In such
instances surface mounted joints are finding much favour in the UK.
These are bonded to the deck via resin nosings and although one
manufacturer has gained a good track record, other systems, based on
the same principle, have failed during service trials due to inadequate
anchorage.
Deck joints which experience mechanical anchorage failure will
usually require considerable remedial works to return the joint to a
serviceable state. This is equally true of segmental bolted joints
where damage to a single segment results in the need for the
replacement of that component; the time taken for the repair of bolted
joints can be significantly increased, and thus serviceability
seriously impaired, due to the ingress of salt and grit.
The acceptability of maintenance operations, whether they be regular
minor works to APJs or infrequent major repairs to failed mechanical
joints, must be considered with respect to the disruption that would be
caused to road users, particularly on major arterial routes such as the
M25. Such considerations may demand the selection of a more reliable
form of joint even if this incurs an increase in the initial capital
cost.
An over-riding factor in the selection of the form of joint may be
the resurfacing requirements for the main carriageway. This is
particularly applicable to multi-span simply supported structures such
as the elevated sections of the M5 and M6 through Birmingham. There
may be a case for re-articulation of the deck and the introduction of
continuity when major maintenance of such structures is carried out.
G. Maunsell & Partners adopted this approach for the refurbishment of
the A19 Tees Viaduct and the subject is to be covered in greater detail
by another author. Such refurbishment exercises can modify the
performance requirements for expansion joints, eg. the transfer of

braking loads between spans. These must be taken into account during
the selection and specification procedure.
The study has shown that up to 70% of the APJs and approximately 6 5 %
of the REJs inspected were leaking. In contrast, leakage through EMR
joints occurred in only 40% of cases and the majority of these were
related to the use of preformed sheet waterproofing systems; the use of
a liquid membrane system would probably alleviate many of these
cases. The expenditure on the repair and rehabilitation of the DoT
stock of concrete bridges has been reported by Wall bank to be in the
order of £60 million per annum, a significant proportion of which has
been necessitated by the effects of chloride laden run-off water
leaking through expansion joints. In cases where concrete sub
structures cannot be provided with protection, e.g. waterproofing the
bearing shelf area and deck ends, then a heavy onus should be placed
upon the performance of the selected joint with respect to
watertightness.
3 Detailing and specifications
The selection of an appropriate joint must be supported by the

conscientious detailing of the joint, drainage systems and the
underlying substructure in order to ensure that a durable design is
achieved.
An Advice Note on Design for Durability, currently being prepared by
the DoT, Holland (1993), contains guidance on the provision of abutment
galleries where conventional abutments are to be used as well as other
design aspects pertinent to expansion joints.
The current UK DoT Specification for Highway Works restricts the
level of detail that can be incorporated in specifications for
expansion joints. Provision is made for the specification of sub
surface drainage and it is strongly recommended that such drainage be
provided where water is likely to pond on the bridge deck adjacent to
the joint.
A large majority of the problems observed with expansion joints can
be attributed to the inefficient drainage of sub-surface water.
Unfortunately the drainage systems currently used either have
insufficient capacity or appear to be unmaintainable and become blocked
over an unacceptably short time scale. Transverse drainage systems can
be detailed such that they can be rodded or jetted and such detailing
should be encouraged wherever possible.
If feasible the inclusion of dual drainage systems should be
considered; transverse drainage can be placed immediately adjacent to
the joint and combined with through-deck drainage some distance away
from the joint. Consideration must, however, be given to the discharge
from such drainage to ensure that other areas of the structure, e.g the
internal surfaces of box decks, are not inadvertently affected by run
off water or put at potential risk.
There is a real need for further developments in the design of sub
surface drainage systems and this will become increasingly necessary
with the introduction of porus asphalt surfacing. The need for such
8 Johnson
drainage can however be alleviated by the use of buried joints if

articulation can be arranged such that the lowest end of the deck also
acts as the point of fixity and that dynamic rotational movements are
1imited.
Despite the restrictions on the engineer's freedom to specify joint
details, it is strongly recommended that the requirements of the
Departmental Standard and Specification are clearly shown on tender and
contract drawings. The adoption of such practices partially resolves
the problem of expansion joints being regarded as secondary components
in bridge design and should prompt both the design engineer and
contractor to consider the detailing and installation of joints more
fully.
The previous discussion of the parameters that should be taken into
account when selecting an appropriate form of joint highlight some of
the deficiencies in the existing specifications. It has been stated
earlier that the engineer is currently unable to apply the full
selection process when either specifying the form of joint to be used
or when considering the contractor's selection of a joint. This state
of affairs may be partially redressed by the introduction of whole life
costing by the Design for Durability Advice Note. It has, however,
been suggested that there may be insufficient cost data available to
apply the method as comprehensively as may be desired, Leeming (1993).
The current study has certainly shown this scarcity of data to be true
of maintenance costs and of estimates of maintenace-free lives for
expansion joints.
4 Concluding remarks
It has been found that the majority of expansion joints have some form
of defect and that the problems identified by previous workers still
persist.
A significant proportion of expansion joints leak and thus represent
a major cause of deterioration of highway bridges.
There has been a notable increase in the use of asphaltic plug
joints throughout the road network in the UK. However, the response of
binders to dynamic and sudden thermal movements remains questionable
and research to define both the required properties and to provide
further guidance to engineers is required.
The current UK DoT Specification does not allow the full selection
process for joints to be implemented by the specifying engineer and
thus, at present, may lead to the installation of expansion joints
which do not provide the best solution for the specific site being
considered. There is, therefore, the need for a revision to the
specification either to enable the engineer to select appropriate joint
types or to comprehensively incorporate the selection procedure in a
form of performance specification. If the latter course is to be
taken, definitions of exactly what constitutes a failed joint will have
to be prepared; these do not exist at present.
There is an increasing amount of interest in the adoption of
integral bridges as a common design concept. There is, however, debate
as to whether deterioration occurs in the surfacing at the interface
between the road pavement and the bridge structure and whether an
expansion joint should be provided at this location. This
deterioration will probably be exacerbated by the introduction of
porous asphalt on to the UK road network. The requirements for such a
joint will be similar to those applicable to a conventional abutment
and the inadequate selection, detailing and specification of such
joints will result in expenditure on remedial works over an
unacceptably short maintenance cycle.
Examples of expansion joints which exhibit satisfactory in-service
performance have been observed during this study and prove that close
attention to detail at the design and installation stages can avoid the
defects which are commonly associated with deck joints.
5 Acknowledgement
This paper is based on work conducted as part of the research programme

for the Transport Research Laboratory, carried out by G. Maunsell &
Partners, and is published by permission of the Director.
The views expressed in this paper are not necessarily either those
of the Transport Research Laboratory nor those of the Department of
Transport.
6 References
Booth, Sweetman and Woolfe Consulting Engineers and Architects (1987).

Bridge Maintenance Survey Results, Report No. 108502. Auckland, New
Zealand
Department of Transport (1989a) Expansion Joints for use in Highway
Bridge Decks. Advice note BA26/88. MMSO, London.
Department of Transport (1989b) Expansion Joints for use in Highway
Bridge Decks. Departmental Standard BD33/88. HMSO, London.
Department of Transport (1991) Manual of Contract Documents for Highway
Works. Vol 1 Specification for Highway Works. HMSO, London.
Holland, D.A. (1993) Design of bridges for durability and maintenance,
in Bridge Management 2, Inspection, Maintenance and Repair (eds
Harding, J.E. Parke, G.A.R. and Ryall M.J.), Thomas Telford Ltd,
University of Surrey, pp889-896.
Johnson, I.D. and McAndrew, S.P. (1993) Research into the Condition and
Performance of Bridge Deck Expansion Joints. Project Report 9
E434A/BC. TRL, Crowthorne.
Koster, W . (1986) The Principle of Elasticity for Expansion Joints,
Mageba SA, Bulach (submitted for 2nd world congress on joints and
bearings, San Antonio).
Lemming, M . (1993) The application of life cycle costing to bridges, in
Bridge Management 2, Inspection, Maintenance and Repair (eds
Harding, J.E. Parke, G.A.R. and Ryall M.J.), Thomas Telford Ltd,
University of Surrey, pp574-583.
PIARC (1987) Expansion joints, Road Bridges Technical committee report
No. 11, in XVIII World Road Congress (eds Drabble, G.M.W. and
Mallett, G.P.), J.B. Shears and Sons, Brussels, pp 33-44.
10 Johnson
Price, A.R. (1984) The Performance In-service of Bridge Deck Expansion

Joints, TRRL Report No. LR1104, TRL, Crowthorne.
Wall bank, E.J. (1989) The Performance of Concrete in Bridges, a Survey
of 200 Highway Bridges. HMSO, London.
Weisgerber, F.E. Minkarah, I.A. and Malon, S.R. (1987) Observations
from a field study of expansion joint seals in bridges. Transp.
Res. R e c , 1118, 39-42.
2 R E S E A R C H REQUIREMENT S
F O R BRIDG E DEC K
E X P A N S I O N JOINT S
J.R. C U N I N G H A ME
Transport Research Laboratory, Crowthorne, UK
Abstract
Any typ e o f join t i s a potential sourc e o f weakness .
Modern concret e an d stee l bridge s hav e joint s t o
accommodate th e movement s o f th e bridg e dec k relativ e t o
the adjacen t pavement . Man y o f thes e joint s leak ,
allowing wate r an d chloride s t o penetrat e dec k ends ,
abutments, pier s an d crosshead s an d caus e corrosio n o f
the reinforcin g steel . The performanc e o f curren t joint s
is ver y variable . Som e ar e satisfactory , bu t many hav e
service live s o f onl y a fe w years an d ar e expensiv e t o
maintain.
Joint performanc e i s determined b y a comple x
combination o f factors . Som e o f th e main factor s are ; th e
movements a t th e join t resultin g fro m therma l variation s
and whee l loading , th e propertie s o f th e materials used ,
and th e installatio n condition s an d standar d o f
workmanship.
This pape r give s a n overvie w o f researc h t o develo p a
performance base d specificatio n base d o n approva l
testing. Laborator y test s ar e neede d t o simulat e
movements an d whee l loadin g o n ne w joints , and materia l
tests nee d t o b e develope d fo r us e i n quality control .
Keywords: Bridge , Expansio n joints . Surveys , Fatigu e
testing.
1 Introduction
Bridge substructure s mov e ver y slowly , bu t th e dec k i s

subject t o shor t duratio n traffi c induce d movements , a s
well a s thos e du e t o change s i n the environmen t suc h a s
temperature. Th e amoun t o f movemen t depend s o n th e span ,
traffic loadin g an d for m o f construction . An y typ e o f
joint i n a structur e i s a potential sourc e o f weaknes s
and expansio n joint s ar e n o exception .
Designing ou t som e o r al l o f th e expansio n joint s ma y
be appropriat e fo r ne w bridge s an d som e existin g
multi-span structures , bu t fo r th e grea t majorit y o f
existing highwa y bridges expansio n joint s wil l hav e t o b e
© Crow n copyright
12 Cuninghame
maintained, an d replace d severa l time s durin g th e lif e o f

the structure . Irrespectiv e o f th e typ e o f join t used ,
the requirement s ar e simpl y stated .
• T o provid e a sea l betwee n tw o part s o f th e structur e

which ar e movin g relativ e t o on e another .
• T o provid e a smooth , quiet , durable runnin g surfac e
for traffic .
2 Joint types
Several type s o f expansio n join t ar e use d i n the U K t o

cater fo r differen t movemen t ranges ; buried (20m m ma x
movement), asphalti c plu g (40mm) , nosin g (40mm) ,
elastomeric, i n metal runner s an d cantileve r com b o r
tooth (mechanica l joints) . Th e maximu m capacit y o f th e
latter thre e type s varie s accordin g t o manufacturer an d
type. Requirement s fo r expansio n joint s o n highwa y
bridges ar e give n i n Departmental standar d B D 33/8 8
(Department o f Transport , 1989a) , wit h additiona l advic e
in a companio n Advic e Note , BA 2 6/88 (Departmen t o f
Transport, 1989b) .
In genera l ther e ar e fe w problem s wit h elastomeri c
joints. Th e elastomeri c sea l may b e damage d b y shar p
debris, bu t i t ca n b e replaced . However , th e majority o f
the U K bridg e stoc k consist s o f shor t spa n bridges ,
asphaltic joint s ar e more commo n tha n th e large r capacit y
types. Burie d an d nosin g joint s wer e widel y use d i n th e
1910's, bu t asphalti c plu g joint s ar e currentl y th e mos t
popular type . Curren t researc h a t th e Transpor t Researc h
Laboratory (TRL ) i s therefore concentrate d o n asphalti c
plug joint s (bu t larg e mechanical joint s may becom e mor e
widely use d i n futur e t o provid e a singl e join t i n a n
otherwise continuou s structure .
3 Review of Department of Transport (DOT) research
Trials an d survey s o f severa l join t type s were carrie d

out i n th e earl y eighties . These include d burie d joint s
(Price, 1 9 8 2 ) , ope n ga p (nosing ) typ e joint s (Price ,
1983) an d a seve n yea r stud y coverin g al l th e main type s
then i n servic e (Price , 1984) . Sinc e then , burie d joint s
and ope n ga p (nosing ) joint s hav e largel y bee n supersede d
by asphalti c plu g joint s an d thes e hav e bee n furthe r
developed wit h th e introductio n o f ne w materials .
Nevertheless, many o f th e conclusion s o f tha t stud y ar e
still valid . The y ca n b e summarise d a s follows .
• Join t performanc e i s determined b y a comple x

combination o f factor s whic h var y wit h an d withi n
joint types .
Research into bridge deck expansion joints 1 3
• A principal facto r affectin g join t performanc e i s the

nature an d magnitude o f the structural movement s at
the join t resultin g fro m therma l variation s and
traffic loading .
• Installatio n condition s ca n severely limi t the life of
a joint ; poor workmanshi p was responsible fo r many
premature failures .
• Th e choice of materials ca n significantly affec t lon g
term performance .
• Structure s wit h relativel y flexibl e deck s can present
severe condition s fo r certain type s of joint. Joint s
are als o affecte d b y bearing performance .
• Weaknesse s i n joint desig n were responsibl e for
premature failure , mainly i n the wheel tracks .
• Leakag e a t joints ofte n cause d damag e to substructure s
and bearings .
The emphasi s was then switche d t o laboratory testing .

Nottingham Universit y wer e commissione d t o develop a
laboratory tes t facilit y fo r movement testin g o f buried
joints. Join t specimen s wer e mad e up in a tray, 0.5 m wide
by 1.0 m long. The tray i s in two halves, connected onl y
by the joint material . Horizonta l an d vertical movement s
can be applied separatel y o r in combination and
rotational movemen t ca n be applied i n place of the
vertical movemen t i f required. Bot h rapi d movement s due
to traffi c an d slow change s simulatin g therma l variation s
can be applied. Th e specimen i s housed i n a temperatur e
controlled chamber .
Tests wer e carrie d ou t on buried joint s (Brow n et al,
1991) to investigate therma l movement , traffi c induce d
horizontal, vertica l an d rotational movement , and the
effects of test temperatur e an d stress conditio n of the
joint at the start of the test. It was found tha t typica l
service failure s coul d b e reproduced i n the laboratory
and the performance o f the test specimen s was comparable
with observation s o f service performanc e by Price (1982) .
This facilit y has now been installe d a t TRL, updated and
developed t o test asphalti c plu g joints , see Figure 1.
At the same time , Reading Universit y wer e aske d to
assess th e feasibility o f a laboratory tes t facilit y
capable o f applying servic e whee l load s i n the
laboratory. Man y possibl e method s o f applying whee l
loading wer e considered . Simila r equipmen t i s used in
pavement tes t facilities , but none of the existing
designs was suitable. Rotary machine s of manageable siz e
would appl y significan t scrubbin g forces , linea r design s
apply th e load forward s and backwards whic h may well
affect expansio n join t performance , and most hav e too
slow a cycling rate . The final desig n was a rotar y
machine wit h straigh t section s ove r thre e tes t areas.Fo r
reasons o f cost a third scal e machine was envisaged.
However, a parametric stud y by Aberdeen Universit y showe d
14 Cuninghame
that a scal e mode l coul d no t provid e th e basi s o f a n

approval test . Th e visco-elasti c propertie s o f asphalti c
joints woul d mak e i t impossibl e t o scal e th e effect s o f
speed, temperatur e an d materia l propertie s (e g aggregat e
s i z e ) , an d henc e t o relat e th e tes t result s t o behaviou r
at ful l scale . Also, i t would b e uneconomi c t o
manufacture scal e model s o f mechanical joints , and t o b e
sure tha t th e mode l woul d behav e i n the sam e way a s a
full siz e joint . Th e solutio n t o thi s proble m
is outline d i n sectio n 6 below .
4 Research objectives
The curren t DO T approva l syste m fo r expansio n joint s fo r

highway bridge s require s a on e yea r tria l installatio n
period o n a highway bridge , i e approva l testin g i n th e
field.
Trials canno t provid e reliabl e informatio n o n th e
performance o f a ne w join t quickl y enoug h t o b e useful.
There ar e tw o problems . Condition s var y s o widely tha t
performance o n on e bridg e canno t b e regarde d a s
representative, an d secondl y i t i s ofte n impossibl e t o
identify th e caus e o f an y deterioratio n i n service . Th e
result i s tha t th e Enginee r doe s no t kno w whic h
applications th e join t ma y b e suitabl e for , an d th e join t
manufacturer ha s littl e guidanc e a s t o ho w t o modif y th e
joint t o improv e performance . Ther e i s als o a natura l
reluctance t o instal l untrie d join t type s o n heavil y
trafficked structure s becaus e o f th e possibl e traffi c
disruption. Abou t 1 0 years o f experience , an d repeate d
condition survey s woul d b e require d t o determin e th e
performance o f a join t i n the ful l rang e o f conditions ,
by whic h tim e i t may b e obsolete .
A laborator y base d approva l syste m ha s severa l
advantages. Th e severit y o f th e applie d movemen t an d
loading ca n b e varie d t o asses s a join t fo r us e i n a
variety o f condition s allowin g th e Enginee r t o selec t a n
appropriate join t fo r eac h application . Join t
manufacturers wil l hav e a consistent , testabl e
requirement t o desig n t o an d coul d modif y an d retes t ne w
joints quickl y i n the even t o f earl y failure . Also th e
research necessar y t o defin e a se t o f approva l test s wil l
lead t o a bette r understandin g o f join t behaviou r an d th e
conditions whic h joint s hav e t o withstand i n service . Th e
disadvantage i s o f cours e tha t th e environmen t (weathe r
etc) canno t b e reproduce d a t justifiabl e cos t i n th e
laboratory.
There ar e tw o basi c requirement s whic h a laborator y
system wil l hav e t o meet . Th e tes t must reproduc e th e
types o f join t deterioratio n whic h occu r i n service , an d
it must produc e result s quickl y t o asses s ne w products .
The ai m of current an d future researc h i s to reduce

expenditure o n maintenance an d replacement o f expansio n
joints, an d to prevent damag e to bridge substructure s due
to fault y joints . Specifically , th e objectives ar e as
follows.
• T o determine the performance o f joints currentl y in

service.
• T o develop suitabl e test s fo r type approva l of joints,
and fo r routine qualit y assuranc e testin g o f installe d
joints.
• T o devise a performance base d specificatio n fo r bridge
expansion joints .
• T o develop improve d desig n guidanc e an d promote highe r
standards o f installation .
• T o establish agree d failur e criteri a an d provide
guidance fo r maintenance authorities .
5 Asphaltic plug joints (APJ)
As note d previousl y asphalti c plu g joint s hav e see n

greatly increase d usag e ove r recen t year s and are now the
most commo n type . They ar e relatively chea p and quick to
install, reducin g traffi c managemen t costs . However,
durability i s a problem, wit h som e joint s failin g i n a
matter o f months. The most urgen t nee d i s therefore to
understand an d improve the performance o f this typ e of
joint. A typica l APJ is shown i n Figure 2.
Many bridge s ar e resurfaced afte r abou t seve n year s
and i t is convenient t o replace the expansion joint s at
the sam e time . It is calculated tha t APJ's are economic
if the y hav e a service lif e of at least 5 years. Thi s
gives a target lif e of 5-7 years.
The performanc e o f an APJ depends on:
• Th e material propertie s o f the plug.

• Th e installation (procedur e an d standard of
workmanship).
• Th e design o f the joint, eg joint geometry .
• Th e movements occurrin g i n service (therma l and
traffic induced .
• Th e wheel loadin g impose d by heavy good s vehicle s
(HGVs).
All o f these may vary widely . Experienc e suggest s tha t

the mos t importan t factor s are the actual materia l
properties o f the installed join t (a s opposed t o the
design values ) an d the standard o f installation an d to a
lesser exten t the movements an d wheel loads .
Failure o f bridge bearing s ca n also affec t the
16 Cuninghame
expansion joints . Seize d bearing s wil l alte r th e positio n

of th e centr e o f rotatio n o f th e dec k unde r a vehicl e
road an d caus e arching' o f th e dec k instea d o f linea r
x
thermal movement . A partiall y seize d bearin g wil l ten d

to mov e i n a stick-slip' manner, an d th e resultin g rapi d
A
movements wil l b e more damagin g t o th e expansio n joint .
5.1 Definition of failure

The mai n fault s whic h occu r i n APJs ar e leakage ,
tracking, crackin g an d debonding . Th e detail s o f th e
faults foun d b y surve y an d thei r cause s ar e discusse d b y
Johnson an d McAndre w (1993) . However , ther e i s no agree d
definition o f join t failur e an d maintenance practice s
vary. On e o f th e problem s fo r durabilit y testin g i s tha t
failures ma y b e progressiv e an d interactive .
There ar e tw o view s abou t whether leakag e constitute s
failure o f th e joint . On e i s tha t mos t joint s lea k s o th e
designer shoul d mak e provisio n fo r water t o b e draine d
away harmlessly . Th e othe r vie w is , that i t i s th e
function o f th e expansio n join t t o preven t wate r gettin g
to th e structur e belo w th e roadway , therefor e a leakin g
joint ha s failed . A s usua l ther e i s no simpl e answer . Fo r
new design s i t i s good practic e t o provid e drainag e an d
provision fo r cleanin g th e drainag e channels , but fo r
some existin g bridge s thi s i s not possible . Th e
definition o f join t failur e shoul d includ e a n assessmen t
of th e consequence s o f th e fault . Where a leakin g join t
would allo w wate r t o reac h part s o f th e structur e
susceptible t o chlorid e damage , the n a leakin g join t
should b e deeme d t o hav e failed .
A certai n amoun t o f trackin g ca n occu r withou t
affecting th e performanc e o f th e joint . However, seriou s
tracking ca n lea d t o impac t load s which i n turn caus e
deterioration i n rid e qualit y an d damag e t o th e join t an d
adjacent surfacing . A limi t o n th e acceptabl e surfac e
profile acros s th e joint , e g dept h o f tracking , woul d
seem t o b e required . I t may b e possibl e t o bas e thi s
limit o n th e standard s applie d t o roa d pavements .
Cracking o f th e join t materia l an d de-bondin g fro m th e
adjacent surfacin g ar e progressiv e failure s whic h ar e
likely t o lea d t o leakag e an d potholing . Surfac e crack s
may b e detecte d b y visua l examinatio n an d shoul d b e
repaired t o preven t furthe r deterioration . However , a
joint whic h appear s perfectl y soun d o n th e surfac e ma y
leak du e t o subsurfac e faults . Water ma y collec t i n th e
adjacent surfacing , an d traffi c loadin g ca n generat e
sufficient hydrauli c pressur e t o forc e th e water pas t an y
weak point .
5.2 Installation
Several factor s actin g separatel y o r i n combinatio n ca n
have a ver y larg e effec t o n th e servic e performanc e o f
the joint . Som e o f th e main one s are :
• Th e condition o f the prepared join t area . If the

adjacent surfacin g i s wet, it can be difficult to dry
the side s of the trench an d achieve a good bon d
between the joint and surfacing.
• Th e cost and inconvenience o f delays to traffic are
becoming increasingl y importan t an d sometimes lea d to
great pressur e o n the joint installe r to complete the
work quickl y (perhap s coverin g up unexpected problem s
instead o f taking tim e to solve t h e m ) . Re-openin g the
carriageway t o traffic befor e the joint has cooled
properly ha s also cause d trackin g i n some joints .
• Obtainin g a watertight sea l betwee n the joint and the
existing dec k waterproofin g membran e i s difficult, and
may wel l explai n why many apparentl y soun d joint s
appear to leak. Thi s proble m i s exacerbated b y poor
drainage allowin g pondin g an d build-up o f hydraulic
pressure i n the surfacing adjacen t to the joint.
Good preparation , attentio n to detail and adherence to

the manufacturer s procedur e ar e all necessary t o achieve
a durabl e joint . The means to achieve thi s are education,
adequate supervisio n b y the client, and QA testing. An
example o f the benefit o f testing the finished produc t
can be drawn fro m anothe r are a o f bridge construction . It
is well know n i n steel fabricatio n tha t a better wel d is
produced i f the operator know s i t is going to be X-rayed.
The Count y Surveyor s Societ y projec t mentione d belo w
will ensur e tha t the views of client authoritie s ar e take
into account . The detailed installatio n procedur e mus t
remain a matter fo r the manufacturer, but guidance ca n be
given to promote goo d practice , in for example:
• Mi x proportions an d laying procedur e - the volume and

temperature o f both aggregat e an d binder an d the
laying an d compaction procedure .
• Environmenta l restrictions , eg temperature and
humidity.
• Th e surface characteristic s o f the surfacing an d deck,
particularly dryness , roughness an d the effectiveness
of priming .
• Th e width an d fixing o f the cover plat e and the
characteristics an d area of debonding.
6 TRL research
The Departmen t o f Transport hav e commissione d a programm e

of researc h a t TRL to meet the objectives outline d in
section 4 above. The approach to the problem wil l be
mainly experimental , the aim is to establis h
relationships between: -
(a) servic e performance ,
(b) larg e scal e laborator y tests .
18 Cuninghame
(c) type approva l test s fo r ne w joint s

(d) routin e qualit y assuranc e test s o n sample s o f
as-installed join t material .
6.1 Service performance

The result s o f a n extensiv e surve y wer e reporte d b y
Johnson an d McAndre w (1993 ) an d furthe r dat a o n servic e
performance wil l b e adde d b y a project se t up b y th e
County Surveyor s Societ y Highwa y Researc h Group . Severa l
other survey s hav e bee n undertake n b y Loca l Authorities .
These wil l hel p t o establis h th e cause s o f failure , th e
maintenance practice s o f bridg e authorities , an d th e
costs o f join t maintenance . Additiona l informatio n an d
case studie s ar e obtaine d fro m sit e visit s an d
discussions wit h join t installer s an d maintenanc e
authorities.
6.2 Large scale testing

Movement test s ar e i n progress usin g th e tes t ri g
developed fro m th e Nottingha m Universit y equipment .
Initial indication s ar e tha t a well mad e an d correctl y
installed AP J i s no t affecte d b y slo w ^thermal ' movement s
at temperature s o f 20° C an d above . However th e numbe r o f
cycles o f movemen t whic h ca n b e sustaine d fall s rapidl y
for temperature s belo w 0°C . This confirm s anecdota l
evidence tha t crack s ofte n appea r afte r a shar p frost .
A ful l scal e linea r whee l loadin g ri g ha s bee n
designed a t TR L an d i s currentl y bein g manufactured . A
half-axle whee l loa d o f u p t o 6 tonnes wil l travers e a
1.5m lon g tes t sectio n a t speed s o f u p t o 30kph , a t a
repetition rat e o f 2 5 passe s pe r minute. Th e wheel i s
raised o n th e retur n pas s s o tha t th e wheel run s ove r th e
joint i n on e directio n only . Th e ke y t o thi s machine i s a
spring/mass resonan t drivin g mechanis m whic h deal s wit h
the hig h acceleration s a t eac h en d o f th e stroke , an d
reduces th e energ y requiremen t t o manageable proportions .
It i s hoped t o begi n testin g earl y i n 1994 . This ri g wil l
also b e capabl e o f testin g elastomeri c an d larg e
mechanical (modular ) joints .
6.3 Type approval tests

These test s wil l b e develope d fro m th e larg e scal e
testing. I t i s hoped eventually , t o combin e th e movemen t
and wheelloadin g equipmen t s o tha t onl y a singl e tes t
would b e require d o n a ne w joint , with a resultan t cos t
saving t o th e manufacturer .
The tes t loa d an d enduranc e requirement s coul d b e
derived fro m existin g desig n whee l loa d spectra , fo r
example i n tabl e 1 1 o f th e U K bridg e desig n code , BS 540 0
part 1 0 (BSI , 1980) . Europea n traffi c loadin g standard s
are als o bein g developed . Severa l spectr a ma y b e
required t o cove r differen t classe s o f road . Give n th e
endurance unde r constan t amplitud e loading , an d a desig n
spectrum, a fatigue assessmen t mode l ca n be used to

relate enduranc e i n the laboratory tes t to service lif e
under traffi c loading . Thi s aspec t applie s to mechanical
joints whic h may be susceptible to fatigue cracking .
Such a test (o r set of tests) woul d b e relatively
expensive, an d so would likel y be a once onl y tes t to
establish th e performance o f a new design of joint.
6.4 Material tests

A se t of static an d dynamic test s are being develope d to
establish th e properties o f a sample of APJ material
which ca n be related t o the properties o f the specimens
used i n the large scal e tests . Potentially suitabl e test s
used fo r pavement material s ar e also bein g investigated .
A 150m m cub e mad e up in a wooden moul d i s being use d
for thes e test s (rathe r lik e a concrete c u b e ) . Thi s typ e
of sampl e woul d b e relatively eas y to produce on site and
so coul d for m the basis of a quality contro l test . It is
hoped tha t simpl e materia l test s coul d be carried out
using equipmen t commonl y availabl e i n Local Authorit y or
commercial laboratories . The material propertie s likel y
to affec t APJ behaviour ar e the characteristics o f the
binder, aggregat e typ e and grading (whic h may be
patented), an d mix proportions.
7 Conclusions
Most expansio n joint s currentl y i n service do not prevent

the ingres s of water an d chlorides to parts of the bridge
structure wher e the y ca n cause significan t damage .
Asphaltic plu g joint s are cheaper tha n othe r type s and
may perfor m ver y wel l i n service. A servic e lif e of 5-7
years i s desirable, but the performance o f joint s
currently i n service i s highly variabl e an d some fai l
within tw o years. This has led to high maintenanc e costs ,
and a programme o f research i s being se t up to address
this problem .
The objective s o f the research ar e to identify the
main factor s whic h determin e servic e performanc e and
develop a performance base d specification . Thi s will be
based o n two levels of testing; typ e approva l testin g to
supplement o r replace the current tria l perio d fo r the
introduction o f new joints, and quality assuranc e testin g
for use in joint installation .
8 Acknowledgements
Crown Copyrigh t 1993 . The views expresse d i n this pape r

are not necessarily thos e of the Department o f Transport.
The contributio n o f Mr R W Jordan to the research
described i n this pape r i s gratefully acknowledged .
20 Cuninghame
9 References
Brown S F , Broderic k B V an d Sha m K C (1991) . Enduranc e

testing o f bridg e dec k expansio n joints . TR L
Contractor Repor t CR297 .
British Standard s Institutio n (1980) . Steel , concrete an d
composite bridges . Part 1 0 cod e o f practic e fo r
fatigue. Britis h Standar d B S 5400 . Britis h Standard s
Institution, London .
Department o f Transpor t (1989a) . Expansio n joint s fo r us e
in highway bridg e decks . Departmental Standar d B D
33/88. Departmen t o f Transport , London .
Department o f Transpor t (1989b) . Expansio n joint s fo r us e
in highway bridg e decks . Departmental Advic e not e B A
26/88. Departmen t o f Transport , London .
Johnson I D an d McAndre w S P (1993) . Researc h int o th e
condition an d performanc e o f bridg e dec k expansio n
joints. Transpor t Researc h Laboratory , TR L repor t no .
PR9. Transpor t Researc h Laboratory , Crowthorne .
Price A R (1982) . Trial s o f burie d joint s an d surfacing s
on a composit e motorwa y viaduct . Departmen t o f
Transport, TR L repor t LR1023 , Transport Researc h
Laboratory, Crowthorne .
Price A R (1983) . Th e performanc e o f nosin g typ e bridg e
deck expansio n joints . Department o f Transport , TR L
report LR1071 , Transport Researc h Laboratory ,
Crowthorne.
Price A R (1984) . Th e performanc e i n servic e o f bridg e
deck expansio n joints . Department o f Transport , TR L
report LR1104 , Transport Researc h Laboratory ,
Crowthorne.
3 B E A R I N G PERFORMANC E
J. LUN D
Allott & Lomax, Manchester , UK
Abstract
This pape r outline s th e function s o f bearing s an d review s
types o f bearings . Bearin g relate d fault s ar e given , bot h
those independen t o f th e typ e o f bearin g an d thos e
associated wit h specifi c type s o f bearing . Acces s problem s
are give n an d th e effect s o f bridg e dec k continuit y ar e
discussed.
Keywords: Bearin g Functions , Bearing Types , Bearin g
Faults, Access , Continuity .
1 Introduction
In th e U K th e desig n lif e o f bridge s i s taken t o b e 12 0

years. Thi s doe s no t mea n tha t it s actua l lif e i s 12 0
years althoug h i t i s ofte n take n t o b e a typica l lif e
expectancy despit e man y bridge s havin g significantl y
longer o r shorte r lives . To remai n serviceabl e throughou t
its lif e a bridg e wil l occasionall y requir e maintenanc e
and repai r an d thi s include s replacemen t o f component s
such a s bearings , joint s etc . I t shoul d no t b e expecte d
that component s wil l endur e fo r th e lif e o f th e bridg e
itself. Som e component s ma y b e expecte d t o las t longe r
than others , fo r exampl e bearing s ar e expecte d t o las t
longer tha n joints , but th e impac t o f replacemen t i s mor e
severe fo r som e component s tha n other s an d generall y
bearing replacemen t ma y b e considere d t o b e on e o f th e
most onerous . Th e cos t o f bearin g replacemen t ma y b e a
factor man y time s th e cos t o f th e bearin g itsel f involvin g
such expensiv e item s a s traffi c diversions , temporar y
support t o th e superstructure , jackin g facilitie s etc .
It i s essentia l tha t th e desig n o f a structur e no t
only consider s th e functio n o f th e bearings , bu t als o
considers th e lif e expectanc y o f th e bearings , acces s fo r
inspection an d maintenance , an d eas e o f possibl e
replacement. Th e for m o f structur e shoul d als o b e
considered wit h regar d t o provisio n o f bearing s an d
joints. A continuou s bridg e dec k fo r multi-span bridge s
Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published in 199 4

by Taylo r & Francis, 2 Park Square, Milton Park, Abingdon, Oxon, OX1 4 4RN. ISBN : 0 419 1903 0 9 .
22 Lund
may significantl y reduc e th e numbe r o f bearing s o r remov e

them altogethe r a t intermediat e piers , and th e s o calle d
"integral bridges " may remov e th e nee d fo r bearing s a t en d
supports.
2 Bearing function and types
2.1 Bearing function

The principa l functio n o f bearing s i s to transfe r vertica l
and/or horizonta l load s fro m th e bridg e superstructur e t o
the substructure . Howeve r i n addition the y als o hav e t o
accommodate rotation s an d translation s o f th e
superstructure relativ e t o th e substructur e abou t eithe r
one o r bot h o f th e principa l axe s o f th e bearing .
Rotations ma y b e accommodate d b y :
a) Compressio n o f th e bearin g materia l itsel f e.g .

rubber i n eithe r a n elastomeri c bearin g o r i n a
metal po t bearing .
b) Rotatio n o f th e to p sectio n relativ e t o th e botto m
section o f bearin g usin g eithe r :
i) a rockin g motio n o n a n interpose d meta l stri p
or poin t e. g meta l rocke r o r rolle r bearing .
ii) a slidin g motio n o n a n interpose d spherica l
sliding plan e e.g . meta l spherica l knuckl e
bearing.
Translation ma y b e accommodate d b y :
a) Slidin g o f th e to p sectio n relativ e t o th e botto m

section o f th e bearin g b y provisio n o f a lo w
friction interfac e e.g . usin g a PTFE/stainles s
steel interfac e (use d i n severa l type s o f bearing) .
b) Shea r o f th e bearin g materia l e.g . rubbe r i n a n
elastomeric bearing .
c) Rollin g e.g . meta l rolle r bearing .
Whilst th e bearin g syste m a s a whole canno t totall y

restrict dec k movements , they contro l wher e th e movement s
may occur . I t i s the locatio n o f fixe d bearing s (n o
translation permitted) , guide d bearing s (translatio n
permitted i n on e direction) , an d fre e bearing s
(translation permitte d i n an y direction ) whic h contro l th e
articulation o f th e bridge , an d thi s i s a secon d functio n
of bearings .
2.2 Bearing types

Prior t o th e 1 9 5 0 s bearing s tende d t o b e fairl y basic . O n
!
shorter span s the y typicall y consiste d o f eithe r 2 plate s

of plai n meta l slidin g bearings , sheet s o f lea d o r bitume n
impregnated fel t o r jus t a mortar bed . O n longe r span s
Bearing performance 23
bearings wer e generall y meta l rocke r o r rolle r bearing s o f

cast iron / meehanite , mild o r hig h yiel d stee l ofte n
fabricated b y a loca l stee l fabricator . I n the perio d
following th e mid t o lat e 1 9 5 0 s specialis t bearin g
f
manufacturers introduce d ne w bearin g type s :
a) Laminate d elastomeri c bearings .

b) PTF E wit h lo w frictio n fo r us e i n the interfac e o f
sliding bearings .
c) Rubbe r po t an d spherica l bearings .
d) Hig h strengt h alloy s i n roller bearings .
e) Fabreeka .
Modern bearing s ma y b e generalise d int o tw o mai n

groups, elastomeri c o r metal, with fabreek a bearing s whic h
were ver y popula r durin g th e 1 9 7 0 s formin g a thir d mino r
f
group, se e tabl e 1 . Som e bearing s ar e a combinatio n o f

elastomer an d meta l e.g . po t bearings , but fo r eas e o f
classification thes e hav e bee n include d wit h meta l
bearings i n tabl e 1 .
Bearing faults
Bearing fault s ma y b e classifie d int o tw o groups , thos e

independent o f th e typ e o f bearin g an d thos e associate d
with a specifi c typ e o f bearing .
3.1 Faults independent of the type of bearing

Faults ma y include :
a) Desig n Fault s Inadequate acces s o r clearanc e
preventing adequat e inspectio n
and maintenance .
b) Constructio n Bearing se t incorrectl y o r
Faults otherwise installe d incorrectly .
Irregular contac t o f adjoinin g
surfaces.
Movement o f bearin g restricte d
by som e misplaced constructio n
material o r pack s etc . lef t i n
place.
c) Maintenanc e Leaking joint s o r blocke d
Faults drainage pipe s resultin g i n th e
bearing bein g continuousl y we t
and subjecte d t o roa d salts .
Accumulation o f detritu s an d
birds nest s aroun d th e bearin g
resulting i n a dam p aggressiv e
environment.
d) Fault s i n th e Cracking.
bedding morta r Missing morta r
Steel shim s lef t i n plac e
creating har d spot s an d causin g
24 Lund
Table 1 Bearin g type s
Type Divisio n Sub- Comments

Division
Elastro- Laminate d
meric
Plain Pad } Restricte d t o smal l
Strip } span s (<10m )
Metal Rolle r Single

Multiple
Rocker Linear } Horizonta l movemen t

Point } a t fre e bearin g
} provide d b y
Knuckle Pin } slidin g interfac e
Leaf } PTFE/stainles s
Cylindrical ) stee l
Spherical }
Pot
Plain slidin g 2 plai n meta l plate s

- n o longe r use d
Fixed pi n } Tak e horizonta l

and guid e }loads onl y
bearings
Fabreeka Horizontal movemen t

of a fre e bearin g
provided b y slidin g
interface o f
PTFE/stainless stee l
Others Bitumen shee t
Lead shee t
no longe r use d
Concrete -
rocker
Cement morta r
Bearing performance 2 5
possible corrosio n wit h

subsequent displacemen t o f
mortar.
Gaps betwee n th e mortar an d th e
bearing.
e) Othe r fault s - Vandalism .
Fire damage .
Impact damage .
Apart fro m th e acces s proble m discusse d later , th e

most significan t faul t i s that o f water leakin g fro m
joints abov e th e bearings , with th e proble m mad e wors e b y
the accumulatio n o f detritu s aroun d th e bearin g holdin g
the water . Th e proble m o f leakin g joint s i s o f cours e
removed a t pier s b y th e us e o f continuou s bridg e decks .
Bearing location s shoul d alway s b e detaile d t o ensur e
good access , an d bearing s positione d s o tha t the y ar e
isolated fro m leakin g joints . They shoul d als o b e detaile d
to enabl e eas e o f replacemen t i n the future .
Many design , constructio n an d maintenanc e fault s ar e
the resul t o f ignoranc e relatin g t o th e importanc e o f
correct detailing , installatio n an d maintenance . Thi s
could perhap s b e improve d b y th e publicatio n o f
information sheet s b y th e Departmen t o f Transpor t givin g
guidance o n goo d practic e an d commo n defects .
3.2 Faults associated with specific types of bearing

Although thes e ar e no t directl y relate d t o dec k
continuity, problem s ma y b e exacerbate d b y leakin g dec k
joints an d fault s ar e briefl y describe d i n 3.2. 1 t o 3.2.4 .
3.2.1 Elastomeric bearings

a) Crackin g o f th e elastomer .
b) Irregula r o r excessiv e bulgin g o r splittin g unde r
vertical loads .
c) Shea r failur e unde r horizonta l loads .
d) Slippag e o f th e bearin g o n it s bed .
e) Irregula r contac t wit h th e tw o bearin g surfaces .
Elastomeric bearing s ar e no t repairabl e an d i f badl y

damaged the y mus t b e replaced . Th e elastomeri c bearin g
itself i s therefor e maintenanc e fre e until replacemen t i s
necessary. Elastomeri c bearing s hav e bee n i n servic e fo r
less tha n 4 0 years an d laminate d elastomeri c bearing s ar e
on th e whol e performin g well , thoug h no t s o plai n
elastomeric bearing s whos e us e i s now limite d t o smal l
spans. Th e lif e o f elastomeri c bearing s canno t b e
ascertained thoug h the y ma y wel l las t upto 10 0 years o r
even longer .
26 Lund
3.2.2 Metal bearings

a) Breakdow n o f th e paintwor k protectiv e system .

b) Deformatio n o f meta l plates .
c) Wea r o n moving surfaces .
d) Indentatio n o f plate s cause d b y th e trackin g o f
roller bearings .
e) Cracking .
f) Over-trave l o f moving parts .
g) Fault s relatin g t o fixing s e.g . missin g bolts .
h) Transi t bolt s no t remove d inhibitin g movement .
i) Slidin g interfac e problem s - se e 3.2.4 .
The mos t commo n faul t i s the breakdow n o f th e

paintwork protectiv e system . Thi s i s made wors e b y th e
problem o f leakin g joint s an d detritu s aroun d th e bearing ,
together wit h acces s problem s preventin g maintenance .
Whilst meta l bearing s ma y b e considere d t o hav e a lif e
expectancy approachin g th e lif e o f th e bridg e the y requir e
regular maintenanc e whic h i f no t provide d (a s i s ofte n th e
case) ca n lea d t o significan t deterioratio n an d a muc h
reduced life .
The durabilit y o f meta l bearing s coul d b e vastl y
improved b y th e us e o f no n corrosiv e metals . Bearings mad e
out o f stainles s steel s would cos t approximatel y fou r t o
five time s thos e o f norma l steel s t o manufacture, bu t
installation shoul d cos t th e sam e an d maintenanc e shoul d
cost less . Howeve r th e numbe r o f bearing s needin g
replacement a t a cos t considerabl y greate r tha n th e cos t
of th e bearing s themselve s woul d b e significantl y reduced .
3.2.3 Fabreeka bearings

a) Breakdow n o f th e paintwork protectiv e syste m t o

metal plates .
b) Slippag e o f th e bearin g o n it s bed .
c) Irregula r contac t wit h th e tw o bearin g surfaces .
d) Slidin g interfac e problem s - se e 3.2.4 .
3.2.4 PTFE/stainless steel sliding interface in bearings

a) Rustin g o f th e metal backin g plat e t o th e stainles s

steel plat e s o warping th e stainles s stee l plate .
b) Dirt , scratching , mino r rustin g o r pain t o n th e
stainless stee l plate .
c) Wea r o f th e PTFE .
d) Over-trave l o f th e slidin g interface .
4 Access problems
The inaccessibilit y o f bearing s create s a major proble m

for thei r inspectio n an d maintenance . Thei r accessibilit y
may b e considere d i n two ways :
a) Th e accessibilit y o f th e bearin g shelf : Bearin g

shelves ar e inherentl y inaccessibl e wit h mos t bearing s
tending t o b e i n suc h location s tha t access*plan t an d
possibly traffi c contro l ar e require d t o ge t t o them .
b) Th e accessibilit y o f th e bearing s o n th e bearin g
shelf: Ofte n i n the pas t littl e consideratio n ha s bee n
paid t o bearin g accessibilit y an d i n many case s the y
have bee n positivel y hidde n fro m view , presumabl y fo r
appearance considerations . I n eve n more cases , usuall y
on concret e bridg e deck s supporte d b y thi n elastomeri c
bearings, a minimum ga p ha s bee n provide d betwee n th e
bearing shel f an d th e dec k soffi t s o tha t th e bearing s
are barel y visibl e le t alon e accessible .
A suitabl e ga p shoul d alway s b e provide d betwee n th e

top o f th e bearin g shel f an d th e dec k soffit , an d a s man y
sides o f th e bearin g shoul d b e accessibl e a s possible . Th e
size o f ga p i s dependent o n suc h detail s a s th e locatio n
of th e bearing s an d th e for m o f dec k constructio n bu t ma y
typically b e i n th e orde r o f 250mm .
If a bearin g canno t b e properl y inspecte d i t i s ver y
unlikely tha t i t will b e adequatel y maintained , a t leas t
until suc h poin t tha t it s deterioration manifest s itsel f
in a proble m elsewher e o n th e bridge . Improve d
accessibility ma y b e provide d a t en d support s b y providin g
an inspectio n galler y typicall y 180 0 mm hig h b y 80 0 m m
wide t o th e rea r o f th e bearings , with acces s t o th e
gallery ofte n gaine d b y a manhole t o th e rea r o r t o th e
side o f th e en d support . Th e inspectio n galler y ha s othe r
benefits, assistin g wit h th e inspectio n o f bridg e dec k
joints, isolatin g th e bearin g shel f fro m leakin g joint s
and providin g acces s fo r othe r maintenanc e work . I n spit e
of th e benefit s presente d b y inspectio n gallerie s ther e i s
still oppositio n t o thei r us e o n an y bridge s b y som e
authorities, whils t other s ar e selectiv e i n their us e
providing the m onl y i n certai n locations . O n bankseat s a n
inspection pi t ma y b e additionall y provide d t o th e fron t
of th e bearin g shel f improvin g acces s t o th e fron t o f th e
bearings.
On continuou s bridg e deck s ther e i s normally onl y on e
line o f bearing s a t eac h intermediat e suppor t an d thi s
automatically improve s acces s t o al l side s o f th e
bearings.
28 Lund
5 The effect of continuity on bearings
The provisio n o f continuou s bridg e deck s ove r intermediat e

supports ha s severa l consequence s o n bearing s give n i n 5. 1
to 5.5 .
5.1 Cost
The numbe r o f bearing s ma y b e significantl y reduce d a t
piers, albei t wit h a consequen t increas e i n siz e o f th e
bearings, resultin g i n a savin g i n both constructio n an d
maintenance costs . Typically tw o bearing s eac h supportin g
a simpl y supporte d dec k t o eac h sid e o f th e pie r ma y b e
replaced b y a singl e large r bearin g supportin g a
continuous deck , bu t thi s i s a n ove r simplificatio n an d
exceptions ma y b e a s follows :
a) Multipl e simpl y supporte d deck s ar e likel y t o

consist o f beams , often a t clos e centres , wherea s
continuous bridge s ma y b e o f differen t for m whic h i s
likely t o requir e fewe r bu t large r bearing s acros s th e
width o f th e deck .
b) Pier s ma y b e constructe d monolithicall y wit h th e
bridge dec k negatin g th e nee d fo r bearing s altogethe r
at th e piers . I n thi s cas e th e pier s ar e subjecte d t o
the longitudina l movement s an d th e rotation s o f th e
deck an d ar e limite d t o situation s wher e th e dec k
movements ar e relativel y smal l o r th e pier s tal l
enough an d flexibl e enoug h t o accommodat e th e
movements. Eve n i n situation s wher e dec k movement s ar e
not necessaril y large , i f the bridg e headroo m i s no t
significantly greate r tha n th e minimum requiremen t
then th e pier s ma y hav e t o b e s o thi n i n orde r t o b e
sufficiently flexibl e tha t the y woul d detrac t fro m th e
appearance o f th e bridge .
c) Wher e dec k continuit y i s provided b y precas t
prestressed concret e beam s lande d o n permanen t
bearings a t th e pier s an d the n made continuou s b y a
narrow insit u concret e crosshea d buil t int o th e end s
of th e beams , the numbe r o f bearing s i s not reduced .
However th e practic e o f erectin g beam s o n th e
permanent bearing s i s not commo n fo r thi s for m o f
construction i n the U K wher e th e beam s hav e tende d t o
be lande d o n temporar y bearing s whic h hav e
subsequently bee n replace d b y a singl e permanen t
bearing locate d beneat h th e insit u concret e crosshead .
5.2 Effect on joints

There ar e n o transvers e joint s betwee n span s o n continuou s
decks an d consequentl y n o leakin g dec k joint s causin g a
serious proble m t o th e conditio n o f th e bearing s beneath .
This i s particularl y importan t o n bearing s containin g
metallic element s susceptibl e t o corrosio n du e t o leakin g
water containin g solubl e roa d salts .
5.3 Accessibility
A singl e lin e o f bearing s o n a pier supportin g a
continuous bridg e dec k give s more spac e fo r acces s fo r
inspection an d maintenanc e tha n tw o line s o f bearing s
supporting separat e bridg e decks .
5 . 4 Pier width
The widt h o f pie r carryin g onl y a singl e lin e o f bearing s
may b e reduce d slightl y fro m th e width supportin g tw o
lines o f bearings . Thi s may b e importan t fo r exampl e i f
it i s necessar y t o locat e a pier withi n a standar d 4 m wid e
central reserv e havin g standar d se t bac k an d clearanc e t o
safety fence s i n fron t o f th e pie r t o U K standards .
5.5 Bearing loads and movements

Bearings supportin g a continuou s bridg e dec k ma y b e
subjected t o totall y differen t loadings , longitudina l
movements an d rotation s tha n bearing s supportin g separat e
simply supporte d deck s o n a n otherwis e simila r multi-spa n
bridge. Th e vertica l loa d o n bearing s a t a pier supportin g
a continuou s dec k wil l b e significantl y greate r tha n thos e
of bearing s supportin g separat e simpl y supporte d deck s
whilst th e rotation s wil l b e significantl y less . Th e
longitudinal movemen t subjecte d t o bearing s wil l als o b e
greater o n a continuou s bridg e tha n o n a bridg e o f
separate simpl y supporte d deck s unles s o n th e latte r th e
decks ar e tied . Bearing s a t en d support s o n continuou s
bridges ar e likel y t o b e subjecte d t o smalle r reactions ,
smaller rotation s an d large r longitudina l movement s tha n
on bridge s o f separat e simpl y supporte d decks . Th e
consequence o f thi s i s that whereas elastomeri c bearing s
may b e viabl e o n a bridg e o f separat e simpl y supporte d
decks the y ma y hav e t o b e replace d b y metal bearing s o n a
continuous bridg e wit h greate r constructio n an d
maintenance cost s offsettin g som e saving s give n i n 5. 1
above. Als o bearin g fault s preventin g th e bearin g
accommodating dec k movement s ma y becom e more significan t
on continuou s bridge s wit h lon g expansion/contractio n
lengths. O n bridge s wit h separat e simpl y supporte d deck s
malfunction o f a bearin g suc h tha t i t does no t accommodat e
deck movement s doe s no t i n general appea r t o b e resultin g
in structura l distres s o f th e bridge . Smal l dec k movement s
associated wit h separat e simpl y supporte d deck s ma y b e
accommodated i n som e othe r manne r whe n no t take n b y th e
bearing, fo r exampl e movemen t o f groun d underlyin g th e
substructure, flexur e withi n th e substructur e etc .
Continuous bridge s howeve r ma y no t b e quit e s o forgivin g
of bearin g malfunction , s o increasin g th e importanc e o f
ensuring th e bearin g perform s a s i t was intended .
30 Lund
6 Summary
a) Ther e appear s t o b e a leve l o f ignoranc e appertainin g

to th e performanc e o f bearing s relatin g t o design ,
construction an d maintenance . On e approac h t o hel p
overcome thi s ma y b e th e productio n o f informatio n sheet s
by th e Departmen t o f Transpor t givin g guidanc e o n goo d
practice an d commo n defects . (Thi s would no t b e limite d
solely t o bearings) .
b) Laminate d elastomeri c bearing s ar e maintenance fre e
and ar e provin g t o b e relativel y durable , but thei r lif e
expectancy i s no t ye t established .
c) Meta l bearing s greates t faul t i s corrosion , an d
consideration shoul d b e give n t o th e us e o f no n corrosiv e
metals, althoug h thi s would increas e th e initia l bearin g
cost b y somethin g i n the orde r o f fou r t o fiv e times .
d) Bearing s ar e ofte n poorl y maintaine d a s a resul t o f
inadequate acces s an d thi s must b e addresse d i n the desig n
of th e bridge . An adequat e ga p must b e lef t betwee n th e
top o f bearin g shel f an d dec k soffit , an d aroun d th e side s
of th e bearings . Inspectio n gallerie s shoul d b e provide d
at en d support s t o provid e acces s fo r inspectio n an d
maintenance o f th e bearing s (a s well a s othe r items ) an d
to isolat e th e bearing s fro m leakin g dec k joints .
e) Bearing s shoul d b e detaile d wit h eas e o f replacemen t
in mind .
f) Continuou s bridg e deck s presen t advantage s relatin g t o
bearings i n tha t th e probabl e reductio n i n the numbe r o f
bearings wil l reduc e constructio n an d maintenanc e costs ,
and eliminat e transvers e joint s a t intermediat e support s
where wate r containin g solubl e roa d salt s may lea k ont o
the bearings . A singl e lin e o f bearing s i s more accessibl e
at intermediat e support s an d th e thicknes s o f intermediat e
supports ma y b e reduced . Continuou s bridg e deck s ar e
likely t o resul t i n a n increas e i n vertical load s an d
horizontal movement s bu t a reductio n i n rotations compare d
to thos e give n b y separat e simpl y supporte d deck s o n a n
otherwise simila r bridge . This will probabl y lea d t o
larger bearing s an d possibl y meta l bearing s rathe r tha n
laminated elastomeri c bearings .
g) Bearing s wit h fault s inhibitin g translatio n t o
accommodate horizonta l dec k movement s d o no t i n genera l
appear t o b e resultin g i n an y manifestatio n o f distres s
elsewhere i n the structur e wher e th e expansion/contractio n
length o f th e bridg e dec k i s small , thus indicatin g th e
movement i s bein g accommodate d i n som e othe r manner . Thi s
may no t b e th e cas e o n th e longe r expansion/contractio n
lengths foun d o n continuou s bridges , so increasin g th e
importance o f ensurin g goo d bearin g performanc e o n suc h
bridges.
7 Reference s
British Standar d 540 0 : Part 9 : 1983 - Bridge Bearing s

Department o f Transpor t Standar d BD20/92-Bridg e Bearings .
Use o f BS540 0 : Part 9 : 1983.
Transport an d Roa d Researc h Laborator y Contrac t Repor t 26 0
Bridge bearin g surve y an d proble m assessment .
4 EFFECT S O F
TEMPERATURE,
SHRINKAGE AN D CREE P
ON INTEGRA L BRIDGE S
B.A. NICHOLSO N
Consulting Engineer, Bicester, UK
Abstract
This pape r give s a n overvie w o f temperature , shrinkag e an d
creep effect s o n bridges . Designer s hav e t o dea l wit h
these effect s i n differen t way s fo r simpl y supporte d spans ,
continuous bridges , an d integra l bridges . Thi s i s one o f
the mai n difference s betwee n th e desig n o f integra l bridge s
and othe r bridges . Temperature , shrinkag e an d cree p al l
give ris e t o effect s whic h ca n onl y b e calculate d wit h a
considerable degre e o f uncertainty , an d consequentl y i t i s
suggested tha t complicate d calculation s ar e normall y
unjustified. Simpl e method s fo r dealin g wit h thes e effect s
in th e desig n o f integra l bridge s ar e suggested , an d
illustrated b y examples .
Keywords; Integra l bridges . Temperature effects .
Shrinkage, Creep .
1 Introductio n
In general , the primar y load s tha t nee d t o b e considere d i n

the desig n o f an y bridg e ar e th e dea d loads , liv e loads ,
and win d loads . Whethe r th e bridg e ha s simpl y supporte d
spans, i s o f continuou s construction , o r i s a n integra l
bridge, th e effect s o f thes e loadin g group s ar e ver y
similar. Th e desig n o f bridge s wit h thes e differen t form s
of articulatio n differ s mainl y i n th e treatmen t o f
temperature an d cree p effects .
In thi s paper , temperature , shrinkag e an d cree p effect s
are describe d i n turn , an d i n eac h cas e th e consequence s
for th e desig n o f integra l bridge s ar e discussed .
Two brie f example s ar e give n t o illustrat e som e o f th e
points raised .
2 Temperatur e effect s
2.1 Temperatur e loadin g

BS 540 0 divide s temperatur e effect s int o tw o categories :
changes i n th e effectiv e temperatur e o f th e bridge , an d

34 Nicholson
differences i n temperatur e throug h th e thicknes s o f th e

deck.
2.2 Temperatur e difference s

Temperature differenc e loadin g cause s a patter n o f interna l
stresses t o for m withi n th e deck . Hambl y (1991 ) gives a
good metho d fo r calculatin g thes e effects , and the y wil l
not b e discusse d furthe r here .
2.3 Effectiv e bridg e temperatur e

Changes i n effectiv e bridg e temperatur e caus e th e dec k t o
expand an d contract . A s thi s i s th e mos t importan t effec t
governing th e desig n o f bridg e joints , and als o ver y
significant i n th e desig n o f bearings , i t i s worthwhile t o
summarise th e wa y i t affect s bridge s wit h differen t type s
of articulation :
A simpl y supporte d bridg e spa n i s allowe d t o expan d an d
contract wit h littl e restraint . I t therefore need s a n
expansion join t a t on e end , an d slidin g bearing s t o allo w
movement ove r tha t support . Temperatur e loadin g doe s no t
create an y longitudina l force s excep t du e t o bearin g
restraint o r friction .
A continuou s bridg e i s als o allowe d t o expan d an d
contract wit h littl e restraint . Assumin g i t to b e fixe d a t
one abutment , i t need s a larg e expansio n join t an d slidin g
bearings a t th e othe r end , an d slidin g bearing s ove r
intermediate supports . Again , temperature loadin g doe s no t
create larg e longitudina l forces .
In a n integra l bridge , expansio n an d contractio n i s
partially restraine d b y th e abutments , which mov e wit h th e
ends o f th e bridge . Th e siz e o f th e longitudina l forc e
induced i n th e dec k depend s o n th e desig n o f th e abutments ,
and thei r interactio n wit h th e soil . N o expansio n joint s
or slidin g bearing s ar e required . However , there i s
relative movemen t betwee n th e abutment s an d th e approac h
road, s o a join t i n th e pavemen t i s needed .
2.4 Integra l bridg e desig n

Expansion an d contractio n o f a n integra l bridg e wil l caus e
the abutmen t t o move . Thi s movemen t wil l b e resiste d b y
earth pressur e behin d th e abutments , and possibl y b y
friction underneath . Abutment s shoul d normall y b e designe d
to minimis e suc h resistanc e t o movement , a s thes e
horizontal force s hav e t o b e carrie d a s compressio n o r
tension b y th e deck . Eve n smal l abutment s ar e likel y t o b e
sufficient t o resis t th e horizonta l traffi c loads .
Granular backfil l shoul d b e use d behin d th e abutments .
The backfil l alway s remain s i n contac t wit h th e abutment s
even a t th e extrem e limit s o f therma l contraction , an d ca n
generate enoug h resistanc e t o horizonta l traffi c load s wit h
only smal l displacements .
There wil l b e therma l movemen t betwee n th e abutment s an d
the approac h road , bu t integra l bridge s ar e likel y t o b e
Effects of temperature, shrinkage and creep 3 5
most usefu l i n circumstances wher e th e pavement doe s not

have t o be interrupte d b y mechanical expansio n joint s eve n
here. I t is suggested tha t asphalti c plu g joint s shoul d b e
used i n the pavement a t the back o f the abutment, or at the
end o f the run-on slab s i f these ar e used. Plu g joint s
work wel l fo r ranges o f movement u p to about 25mm . Simpl e
calculations base d o n the temperatures specifie d i n BS540 0
show tha t a n all stee l bridg e 60 m long , or an all concret e
bridge 85 m long , would hav e a n overall rang e o f therma l
expansion an d contraction o f no more tha n 50m m i n most
locations i n Britain. Thes e bridge s woul d hav e 25m m
movements a t each end . Th e vast majorit y o f bridges hav e
spans n o more tha n these , so integra l constructio n wit h
plug joint s i s widely applicable . Th e plug joint s wil l
need maintenanc e fro m tim e t o time durin g th e lif e o f the
bridge. Base d o n experience i n this countr y an d in the USA
(see NCHR P 3 2 2 ) , this i s not considered t o be a problem .
Integral bridge s shoul d b e designed fo r the sam e
temperature range s a s others bridges . TRR L 69 6 and TRR L
765 sho w tha t th e annual temperatur e rang e i s only abou t
30% les s tha n th e 120 year extreme , so there i s littl e
benefit i n designing fo r a reduced retur n period ,
particularly a s it is normally straightforwar d t o cope wit h
the 12 0 year therma l movement s a s described above .
3 Shrinkag e effect s
3.1 Shrinkag e
Concrete shrink s slightl y a s it ages, an d this ca n affec t
stresses an d deformations i n bridges.
In composit e bridges , the separate component s wil l i n
general shrin k b y different amounts , inducin g a set of
internal stresse s an d curvature o f the deck, an d also a
small overal l shortening . Th e interna l stresse s mus t b e
allowed fo r in the design, an d calculations tak e a simila r
form t o those fo r temperature differenc e loading .
The overal l shortenin g wil l b e most significan t i n an
all concret e bridge . A conservativ e valu e fo r long ter m
shrinkage strai n i n concrete i s about 2 0 0 x l 0 ~ , and abou t
6
half o f this occur s befor e th e concrete ha s reached a n age

of 10 0 days. Eve n fo r an in-sit u integra l bridge , i t is
unlikely tha t integra l behaviou r o f the abutments wil l
start, o r the pavement join t be made, before thi s ag e has
been reached . Thu s a realistic uppe r boun d fo r overal l
shrinkage i s lOOxlO*" , or 10m m for a 100m bridge.
6
3.2 Effec t o n integra l bridge s

The overal l shortenin g o f a concrete bridg e i s much smalle r
than th e range o f therma l movement , and the gradual rat e o f
movement wil l b e swampe d b y the cyclical therma l movements .
Shrinkage i s not significant i n the design o f the
abutments.
36 Nicholson
4 Cree p effect s
4.1 Cree p
Like shrinkage , cree p i s a non-elasti c deformatio n o f
concrete occurrin g ove r time . I n th e lon g term , cree p
increases th e elasti c deformatio n b y a facto r o f abou t
three. I t i s describe d b y a cree p factor , given th e symbo l
0 , an d th e cree p strai n i s define d a s 0 time s th e elasti c
strain. Th e cree p facto r i s tim e dependent , an d normall y
has a lon g ter m valu e o f abou t two , but ca n onl y b e
calculated wit h a considerabl e degre e o f uncertainty .
Creep affect s th e deformation s o f al l bridge s wher e
concrete i s used , bu t it s effec t o n th e stresse s depend s o n
the typ e o f construction .
In a composit e bridge , an y stresse s locke d i n b y
differential shrinkag e betwee n beam s an d th e dec k wil l b e
relieved b y creep . Thi s ca n b e deal t wit h i n design simpl y
by applyin g a reductio n facto r t o th e shrinkag e strain .
BS5400 quote s a reductio n facto r o f 0.43 , whic h assume s
that 0 =2.0.
Creep ha s a significan t effec t i n prestresse d concrete .
In th e cas e o f post-tensione d constructio n an d pre -
tensioned beams , the effec t i s simpl y t o reduc e slightl y
tensioning forc e an d th e force s an d stresse s resultin g fro m
it. Thi s i s taken int o accoun t i n th e desig n simpl y b y
including cree p strai n wit h th e othe r prestres s losses .
The situatio n wher e desig n fo r cree p i s not s o straight -
forward i s where prestresse d concret e i s used, an d th e
stress patter n ha s discontinuities . Thi s occur s whe n
precast beam s ar e use d wit h in-sit u diaphragm s betwee n
spans. Thi s situatio n require s specia l treatment .
Creep ca n hav e beneficia l effects , i n particula r i n al l
concrete integra l bridge s fo r whic h th e rang e o f therma l
expansion ca n b e reduce d b y creep . Thi s effec t wil l
normally b e small , as i n th e exampl e i n sectio n 5.2 .
However, i n case s wher e th e abutment s offe r a hig h degre e
of restraint , force s o n th e abutmen t ma y b e reduce d b y u p
to 3 0 % , but cree p calculation s mus t b e carrie d ou t i f th e
designer want s t o mak e us e o f thi s reduction .
4.2 Desig n usin g prestresse d beam s

This sectio n relate s t o th e desig n o f bot h continuou s an d
integral bridges , as cree p affect s thes e i n th e sam e way .
The difficult y arise s becaus e th e prestres s i s no t
continuous ove r th e supports . Shortenin g o f th e botto m
flange o f th e beam s du e t o cree p cause s th e individua l
beams t o hog , an d open s u p crack s a t th e botto m o f th e
beams ove r th e supports . Th e beam s ca n b e restraine d s o
that thes e crack s d o no t occur , but thi s involve s designin g
for a substantia l saggin g momen t a t th e bea m ends .
It i s interestin g t o compar e th e bendin g momen t diagram s
when hoggin g o f th e beam s i s restraine d (full y continuou s
construction) an d whe n ther e i s n o restrain t (simply -
supported). Figur e 1 shows o n the lef t bendin g momen t

diagrams fo r a simply-supporte d span , made u p fro m th e
combination o f dead an d liv e loading . O n the right ar e the
equivalent diagram s fo r a precast bea m i n a continuou s
structure. Dea d load s ar e carried o n the beam actin g a s
simply-supported, continuit y unde r liv e load s lead s t o
hogging moment s a t the ends. Cree p o f the concrete lead s
to a uniform saggin g moment , includin g ove r th e supports.
Even thoug h thi s analysi s i s simplistic, i t is clear tha t
the tota l bendin g momen t diagra m i s very simila r t o tha t
for th e simply-supported span . Th e creep restrain t momen t
nearly cancel s ou t any benefit fro m continuity .
Creep i s very difficul t t o quantify wit h an y degree o f
certainty, eve n i n the simplest o f designs. I t has als o
been show n tha t ther e i s littl e benefi t t o be obtained fro m
designing th e bridge a s continuous, when a n estimate woul d
have t o be made o f the restraint momen t du e to creep. I t
is therefor e recommende d tha t th e most practica l wa y of
designing a continuous o r integra l bridg e usin g precas t
beams i s to design eac h spa n a s if it was simply-supported .
Soon afte r constructio n ha s finished, whe n cree p ha s onl y
just begun , the restraint momen t wil l no t yet have buil t
up, s o hogging moment s wil l occu r ove r th e supports.
Reinforcement ca n be provided i n the slab t o resist thes e
moments, calculate d ignorin g th e effects o f creep .
Bottom flang e reinforcemen t i n the beams t o allow fo r
the saggin g moment s ove r th e supports i s difficult t o
incorporate int o precas t beams , and has been show n abov e t o
be o f negligibl e benefi t t o the design o f the beam. It s
only purpos e i s to control cracking . I t has been foun d
that crackin g a t the diaphragms ove r pier s i s not a
serviceability problem : th e slab i s not affected, s o ther e
is no penetration o f de-icing salts , and the cracks ar e not
considered t o be unsightly. Ther e i s therefore n o reaso n
to provid e heav y botto m flang e reinforcement , and no reaso n
to attemp t t o complex calculation s t o design it . I t is
recommended tha t som e nomina l botto m flang e reinforcemen t
is provided , bu t that calculation s ar e not normall y
necessary. Som e crackin g shoul d b e expected .
Dead load
L i e load
V
1 11 1 1U Restraint
Total
(a) Simply-supported (b) Continuous

Figure 1 : Bending momen t fo r prestressed beam s
38 Nicholson
5 Example s
5.1 Singl e spa n porta l

Figure 2 show s a schemati c diagra m o f a singl e spa n
integral bridg e wit h ful l heigh t abutments . Th e dec k i s o f
composite construction , wit h stee l beam s an d a concret e
slab, an d span s 30m .
The tota l rang e o f therma l movemen t i s les s tha n 30mm ,
or 15m m a t eac h abutment .
Plug joint s ar e provide d i n th e pavemen t a t th e bac k o f
the abutment .
Creep an d shrinkag e effect s ar e limite d t o restraine d
shrinkage o f th e slab .
There i s n o horizonta l movemen t betwee n th e abutment s
and deck , an d ther e nee d als o b e n o rotation . Bearing s
under th e end s o f th e beam s therefor e serv e n o usefu l
purpose, bu t woul d stil l requir e maintenance . I t would b e
advantageous t o omi t them , an d simpl y buil d end s o f th e
steel beam s int o th e concret e abutments . Thi s woul d als o
apply t o an y intermediat e support s i n longe r bridges , th e
whole structur e bein g monolithic .
Concrete dec k Plug join t
Steel bea m
7 Approach
road
30 m spa n
Figure 2 : Singl e spa n composit e integra l bridg e

5.2 Fou r spa n bridg e

Figure 3 shows a schemati c diagra m o f an integra l bridg e
with ban k sea t abutments , and fou r span s o f 15m , 20m, 20m,
and 15m . Th e deck i s of composit e construction , wit h
precast prestresse d concret e beam s an d a concrete slab .
This exampl e i s based o n Hambly an d Nicholson (1990) ,
which include s a complete se t of design calculations , and
also detaile d calculation s o f creep effects .
Plug joint s ar e provided i n the pavement a t the ends of
the run-o n slabs .
Each individua l spa n ha s been designe d a s if it were
simply-supported, an d Y3 beams hav e bee n used . Nomina l
bottom flang e reinforcemen t o f six 20m m diamete r epox y
coated bar s ha s been provided .
Beams ar e placed o n top of the piers wit h a 200m m gap.
No temporar y support s ar e required. A 400m m thic k in-sit u
diaphragm i s cast aroun d th e beam ends .
Creep wil l caus e th e beams t o hog, and cracking t o occu r
over th e piers. Ther e wil l clearl y b e som e rotatio n o f the
ends o f the beams, so elastomeric pa d bearings hav e bee n
provided.
For a total lengt h o f 70m , the total rang e o f therma l
movement i s about 40mm , o r 20m m a t each abutment . A t the
extreme hig h temperatur e (24 ° above th e mean) the therma l
strain i s about 290/ie , which woul d lea d t o an unrestraine d
movement o f 10m m from th e mean positio n a t each abutment .
An abutmen t movemen t o f 10m m would mobilis e a forc e o f
about 3M N in the case o f the small ban k sea t abutment s use d
in thi s bridge . Thi s 3M N force woul d giv e a unifor m
compressive stres s o f 0. 4 MP a i n the bridge, leadin g t o an
elastic strai n o f 13/ie , which woul d approximatel y doubl e
due t o creep t o abou t 26/ie . Thu s i t can be see n tha t cree p
might reduc e th e temperature expansio n effect s b y abou t
10%.
Run-on sla b In-situ diaphrag m Prestressed beam s
Small
abutment
Flexible pie r
Figure 3 : Four spa n integra l bridg e

designed usin g prestresse d beam s
40 Nicholson
6 Conclusion s
The mai n point s mad e i n thi s pape r are :
a) Th e mai n differenc e betwee n th e desig n processe s fo r a n

integral bridg e an d an y othe r bridg e i s the treatmen t o f
thermal movements . Fo r integra l bridge s th e abutment s
have t o b e designe d t o move .
b) Integra l bridge s shoul d b e designe d fo r th e sam e
temperature range s a s othe r bridges .
c) Asphaltic plu g joint s i n th e pavemen t ar e mos t
appropriate fo r absorbin g th e therma l movemen t betwee n
abutments an d th e approac h roa d fo r th e majorit y o f
integral bridges .
d) N o specia l treatmen t o f shrinkag e effect s i s require d
for integra l bridges .
e) Cree p ha s a significan t effec t o n th e desig n o f
continuous an d integra l bridge s usin g precas t
prestressed beams . An y calculatio n o f cree p effect s ha s
a larg e degre e o f uncertainty , bu t i t i s suggeste d tha t
no calculation s nee d b e attempted . Simpl e desig n rule s
have bee n propose d whic h d o no t requir e calculatio n o f
creep effects , and th e penalt y fo r making thi s
simplification ha s bee n show n t o b e negligible .
f) Cree p ca n reduc e th e effec t o f therma l expansion , bu t
the benefi t i s onl y margina l unles s th e abutmen t offe r a
high degre e o f restraint .
g) As wel l a s expansio n joints , bearings ar e ofte n
unnecessary i n integra l bridges , and ca n als o b e
eliminated.
7 Reference s
BS 540 0 Steel , concrete an d composit e bridges , Britis h

Standards Institution , London .
Emerson, M. (1976 ) Bridge temperature s estimate d fro m th e
shade temperature , TRR L 696 , Transport an d Roa d Researc h
Laboratory, Crowthorne .
Emerson, M. (1977 ) Temperature difference s i n bridges:
basis o f desig n requirements , TRRL 765 , Transport an d
Road Researc h Laboratory , Crowthorne .
Hambly, E.C . an d Nicholson , B.A . (1990 ) Prestressed bea m
integral bridges , Prestressed concret e Association ,
Leicester.
Hambly, E.C . an d Nicholson , B.A . (1990 ) Prestressed bea m
integral bridges . The Structura l Engineer , 6 8 no . 23.
NB: Thi s i s a shortene d versio n o f th e abov e leaflet .
Hambly, E.C . (1991 ) Bridge dec k behaviour , E . & F.N. Spon ,
London.
Oesterle, R.G . Glikin , J.D. an d Larso n S.C . (1989 ) Desig n
of precas t prestresse d bridg e girder s made continuous ,
NCHRP 322 , Transport Researc h Board , Washington, D.C .
R E P O R T O N SESSIO N ON E
A.M. LO W
In hi s presentatio n M r Nicholson ha d propose d tha t i n multi-spa n integra l bridge s

a crac k shoul d be allowe d t o develo p in th e dec k soffi t ove r eac h interna l support .
There was concer n that this crac k would be ove r a rubber bearing where it might rip
the rubbe r skin or it coul d b e just awa y from the bearing where i t might destro y the
shear strength of the section. M r Nicholson suggested tha t double bearing s coul d b e
used, presumabl y with the crac k betwee n them , bu t h e wa s satisfie d wit h the singl e
bearing proposal . I t was suggeste d tha t shrinkag e woul d ten d t o cance l th e cree p
strains causin g th e crack . M r Nicholson's analysi s ha d include d thi s compensatin g
effect.
Mr Nicholson was aske d ho w long an integral bridge coul d be an d he replie d
that i t coul d b e an y length provide d ther e was a sufficien t allowanc e fo r movemen t
between th e en d o f th e dec k an d th e pavement .
The questio n o f the definitio n of the failure o f a road joint was raised. I t was
suggested tha t th e targe t for a joint wa s tha t it shoul d requir e no lan e closure s fo r
inspection o r maintenance between th e resurfacing operations for the road every 15
years.
The integral bridge proposa l was see n as a shifting of the joint proble m fro m
the bridge t o the adjacen t pavement. I t was pointe d ou t tha t this was probabl y very
cost effectiv e becaus e studie s o f actua l maintenanc e cost s ha d show n tha t
maintenance activitie s fo r pavement s coul d be carrie d ou t muc h more cheaply tha n
similar activitie s fo r bridges .
PART TW O
ADVANTAGES
OF BRIDG E
CONTINUITY
5 H I S T O R I C A L REVIE W O F
THE DEVELOPMEN T O F
C O N T I N U I T Y AN D
E X P A N S I O N JOINT S I N
BRIDGES
G.P. T I L L Y
Gifford & Partners Ltd, Southampton, UK
Abstract
This pape r i s concerne d with th e developmen t o f bridg e continuit y an d
methods o f copin g wit h th e expansiv e effect s o f temperature . Earl y type s o f
bridge withou t expansio n joint s wer e abl e t o cop e wit h temperature effects b y
the greate r flexibility o f th e material s an d th e structura l forms the n used.
Modern construction , havin g hig h strength materials , requires articulatio n but i t
has bee n foun d tha t the cos t o f repairin g failed expansio n joint s an d th e
consequental damag e throug h corrosio n is a majo r shortfall. However , recen t
developments an d th e performance s t o dat e o f lon g span jointless bridge s ha s
indicated tha t temperatur e problems ma y have bee n overstate d an d ther e i s
scope fo r reconsiderin g method s o f design .
Keywords: Bridges , Temperatur e Effects, Expansio n Joints, Continuity .
1 Earl y structure s 1600B C - AD1750
The oldes t survivin g structur e in th e U K i s Stonehenge , constructe d i n it s

present for m circ a 160 0 BC using bronze ag e technolog y (Fig.l.) . I t wa s
composed o f 3 0 equa l span s o f 3.14 m having vertica l sarse n stone s o r piers
each weighing abou t 3 0 tonn e an d horizonta l lintel s o f 7 tonne . Th e pier s ha d
top surface s tha t were slightl y dished to provid e a secur e seatin g fo r th e lintel s
and projectin g tenon s t o ke y int o mortic e holes o n th e underside s o f th e lintels .
In additio n th e lintel s were fitte d t o eac h othe r with tongue an d groov e joints .
Stonehenge mus t hav e represente d th e state-of-the-ar t i n it s tim e bu t ther e i s
no evidenc e o f an y othe r jointed structure s of thi s ag e i n th e UK . I t would,
however, b e surprisin g i f n o other s were eve r buil t an d i f ther e were n o
bridges.
Clapper bridge s ar e a mor e primitive for m o f structur e which ma y dat e
from th e bronz e ag e albei t i t ha s been suggeste d tha t they ma y be muc h mor e
recent. On e o f th e bes t known , th e Postbridg e acros s the Dar t in Devon , i s
composed o f thre e 4.6 m simpl y supporte d span s an d i s 2 m wide.
It i s unlikel y that the engineer s o f th e tim e wer e awar e of th e pro s an d con s
of continuit y bu t i n an y cas e thei r structure s were fi t fo r th e intende d purpos e
Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published in 1994

by Taylor & Francis, 2 Park Square, Milton Park, Abingdon, Oxon , OX14 4RN. ISBN: 0 419 19030 9.
46 Tilly
Fig.l. Stonehenge , th e oldes t survivin g lintelle d

structure i n th e UK.
and th e fe w survivor s have stoo d th e tes t o f time .

The Roman s introduce d arc h bridge s t o th e U K albei t th e onl y remnants
are som e substantia l foundation s discovere d a t a number of rive r crossings.
Roman arche s were mainl y o f semi-circula r shape havin g rise s o f n o mor e
than on e thir d span. Th e advantag e of thi s shap e i s tha t all th e stone s i n th e
voussoir ar e identical an d ca n therefore be produce d very accurately. Som e
Roman arche s were buil t withou t an y mortar, a n interesting exampl e bein g
Alcantara bridg e acros s the Tagu s in Spain . Th e tw o centra l arches have span s
of 30 m and th e highes t point s o f th e bridg e ar e 52m above th e river . I n 1812
during th e Peninsula r Wa r one o f th e arche s was destroyed . I t was
subsequently repaire d but th e restorer s were unabl e t o manag e without usin g
mortar.
Some o f th e Roma n viaducts were very long indeed, fo r exampl e th e
aquaduct a t Segovi a i s 728 m long. I t is doubtfu l whethe r an y consideration wa s
given t o th e effect s o f temperatur e and if it had , the engineer s woul d b e
unlikely t o hav e ha d enough confidenc e t o proceed.
The Roma n bridge builder s wer e supporte d b y the advic e o f expert s sent
out b y th e emperor s and forme d int o semi-militar y guilds . The y were
permitted certai n tax allowances whic h wer e se t agains t th e requiremen t to
provide advic e an d follow th e militar y to distan t parts of th e empire . Th e
guilds forme d th e basi s o f professional college s i n which professiona l standard s
were lai d down , Shirle y Smit h (1954) .
When th e Roman s departe d from Britain the state-of-the-ar t fell bac k and
the pointe d arc h was adopted . Thi s require d less skill an d exerte d a lowe r
thrust o n th e abutments . Pointe d arche d were commonl y use d u p t o th e
fifteenth century .
During th e cana l era many arc h structure s were constructe d which ar e still
in operatio n toda y without havin g ha d major strengthening. Thes e structures
were buil t wit h lim e morta r which permitte d relative movement s betwee n th e
blocks. A n exampl e i s show n i n Fig. 2 where a cana l bridge i s collapsin g an d
the block s hav e completel y separated . A bridg e constructe d with moder n
Review of continuity and expansion joints 4 7
stronger an d mor e adhesiv e morta r would have collapse d b y a differen t

mechanism. Thi s bridg e wa s unde r tes t an d reache d a maximu m load o f 211
tonne positione d acros s a lin e a t /3-span, Page (1987) . Th e wea k morta r ha d
1
enabled th e structur e to cop e wit h temperature effects ove r th e year s withou t

any seriou s los s i n loa d carryin g capacity.
Fig.2. Collaps e o f masonr y arc h bridge .
2 Iro n bridges 175 0 - 1890
Iron was introduce d int o bridg e constructio n in th e latte r half o f th e eighteent h

century, Ironbridg e (th e oldes t survivor ) being constructe d across the Sever n a t
Coalbrookdale i n 1777 . Th e earl y structures were a developmen t o f th e arc h
form with no provisio n fo r articulation . Th e end s o f th e rib s were fitte d int o
grooved socket s i n th e abutmen t bearin g casting s an d were considere d t o b e
structurally fixed . Temperatur e effects becam e apparen t at a n earl y stage an d
during constructio n o f Southwar k bridg e i n 181 8 it was foun d tha t th e averag e
effect o f expansio n cause d b y summe r increase s i n temperatur e was a ris e i n
height o f th e crow n o f 38mm . I t seems tha t iron arc h bridge s wer e abl e t o
deflect withou t provisio n fo r expansio n an d without causin g undu e distres s t o
the abutment s i n th e shor t ter m if no t th e lon g term.
The introductio n o f iro n enable d designer s t o tur n to suspende d span s o f a n
altogether mor e flexibl e nature . Notabl e example s tha t are still i n servic e
include Unio n bridg e acros s the Twee d (1820 ) an d th e Mena i bridg e (1826) .
These earl y examples wer e trouble d b y wind induced vibratio n an d ther e wer e
a numbe r o f bridge s tha t were damage d o r destroyed b y wind. Designer s wer e
therefore force d t o focu s thei r attentio n o n improvin g th e dynami c stabilit y o f
48 Tilly
suspension systems an d temperatur e effects wer e o f secondar y importanc e if

they wer e considere d a t all . Jame s Dredge o f Bat h devised a patented syste m
of cabl e configuratio n whic h evidentl y worke d an d severa l o f hi s bridge s ar e
still standing , fo r exampl e Victori a bridge acros s the Avo n a t Bath (1836) .
There wa s n o provisio n fo r expansio n i n Victori a bridge, th e mai n beam s bein g
very lightweigh t an d fixe d int o th e masonr y piers with a mortare d joint. Whe n
inspected recentl y i t was eviden t tha t relative movement s occu r and th e joint s
have converte d fro m a fixe d t o a sliding action.
Robert Stephenson' s Britanni a bridge, constructe d in 1846 , as well a s being
one o f th e mos t researche d designs, wa s a n earl y example o f th e us e o f roller
bearings. Th e bo x girde r is fixe d a t the centra l Britannia tower bu t fre e t o
slide o n cas t iro n rolle r bearings throug h th e othe r towers, Drysdal e Dempse y
(1864). Eac h o f th e hal f length s i s 461m . Usin g a coefficien t o f expansio n
measured b y Professo r Daniel l i n 1831 , it was calculate d tha t the require d
temperature movemen t wa s 152mm . A n expansio n o f 82m m was measure d fo r
a temperatur e increase o f 14°C . Durin g constructio n it was recorde d that
differential temperatur e effects cause d th e bo x t o ben d 57m m laterally an d
63mm vertically .
An awarenes s o f temperatur e effects wa s expresse d i n relatio n t o th e firs t
Tay bridg e (1878) . Th e Boar d of Trad e Inspector , Majo r General Charle s
Scrope Hutchinson , recorde d in hi s acceptanc e report that expansio n effect s i n
hot weathe r shoul d b e reduce d a s muc h a s possible b y painting th e girder s
white. H e als o wrot e tha t he woul d visit th e bridg e agai n t o observ e th e effect s
of a hig h wind when a trai n of carriage s ran across. In the even t th e Ta y bridge
was destroye d b y a gale befor e h e returned.
During th e railwa y er a some 2500 0 bridge s wer e buil t betwee n 183 0 and
1860 an d th e stat e o f bridg e engineerin g wa s develope d ver y considerably. Th e
requirement t o avoi d stee p incline s force d designer s t o build bridges an d
viaducts o f length s an d height s tha t were greate r than anythin g sinc e th e
Roman period . Du e t o th e dynami c effect s o f stea m engines , suspensio n
bridges o f th e tim e wer e to o livel y and designer s turne d to othe r structural
forms, suc h a s throug h trusses , when long spans wer e required . A cas e i n poin t
is th e Hungerfor d bridge , acros s the Thame s i n London , which replace d
Brunei's suspensio n bridge fo r pedestrians . I t was reconstructe d as a 6-spa n
structure incorporatin g th e origina l tower s an d havin g a continuou s truss . Th e
designer apparentl y failed t o mak e du e allowanc e fo r th e tensil e stresse s
developed b y negativ e bendin g ove r th e pier s an d th e to p flange s cracked ; i t i s
not know n whe n the crackin g occurred but i t di d no t appea r to caus e an y
problems becaus e th e bridg e performe d satisfactoril y unti l increase d loadin g
led t o th e nee d fo r extensiv e strengthenin g i n 1979.
Fig.3. Fracture d iro n work over pier of Hungerfor d Bridge.
3 Concret e and stee l 189 0 - 1950
Concrete wa s introduce d towards the en d o f th e nineteent h century , for

example, Glenfinna n Viaduc t which wa s buil t i n 1897 . Th e structur e wa s
composed o f twent y on e span s o f 15 m at heights u p t o 27m . I t was considere d
that ther e was a possibility o f differentia l settlement ; thi s was catere d for by
having 12m m steel plat e slidin g joints a t the crow n of eac h arch . Ther e was n o
provision fo r longitudinal expansion .
The firs t bridg e a t Angel Roa d on th e A406 , buil t i n 1908 , wa s compose d o f
a reinforce d concrete deck o n continuou s beams . I t had 1 7 equal span s an d a
total lengt h o f 223m . Th e structur e evidentl y performe d satisfactorily until
replaced i n 1960.
Wadell (1916) , advise d agains t the unnecessary use o f expansio n joints
because the y coul d weake n th e structure , caus e 'roug h spots' o n th e runnin g
surface, lowe r th e structura l rigidit y and be mor e expensive. Arche s and
suspension bridges di d no t requir e provision fo r expansion a s they coul d
deflect. Howeve r joints would alway s be necessar y when girdere were fixe d t o
masonry. Fo r concrete structures he advise d tha t joints shoul d b e provide d
every 15 m to avoi d cracking.
In Arrol's handbook o f stee l bridge s an d structures , Hunte r (1920) ,
information abou t temperatur e effects o n stee l structure s was provided.
Temperature dat a collecte d ove r the previou s 2 0 years were quote d an d
50 Tilly
measurements o f bridg e expansion s provided . I n a sectio n givin g

recommendations fo r desig n of stee l railwa y bridges, i t was recommende d tha t
provision b e mad e fo r expansio n an d contractio n at the rat e o f 25m m per 30m.
In relatio n t o th e girder s it was state d tha t suitable heav y plate s shoul d b e fixe d
at on e en d an d expansio n bearing s provide d a t the other . I t was als o note d
that ther e shoul d b e restrain t of transvers e motion an d uplift .
Bressey (1933) , th e the n Chie f Enginee r o f th e Road s Departmen t o f th e
Ministry o f Transport , writin g o n bridge s i n general , deal t wit h effects o f
differential settlemen t an d mentione d tha t cast iron arche s could hav e thre e
hinges t o cop e wit h moderate settlemen t an d cite d th e Sunderlan d Wearmout h
bridge whic h wa s abov e ol d min e workings . Stee l bridge s o f th e tim e usuall y
had hinge s a t th e springin g point s bu t thes e wer e designe d t o b e fo r eas e o f
construction. Reinforce d concret e bridges wer e mostl y withou t hinge s but
temporary one s wer e occasionall y use d durin g construction . Expansiv e effect s
due t o temperature s were no t mentioned ,
Bressey note d tha t if multi-span s wer e mad e continuou s ove r th e piers ,
bending moment s woul d be reduce d a t the centr e of th e span s an d less
material would be needed . Th e sam e argumen t would appl y t o singl e span s
fixed t o th e abutments , th e longes t spa n porta l at that time bein g Wisbec h
bridge whic h wa s 28m .
Chettoe an d Adam s (1938 ) writin g abou t desig n of concret e bridges, note d
that th e provisio n fo r expansio n an d contractio n of reinforce d concret e
superstructures i s frequentl y ignored , sometime s wit h unhappy results . Th e
problems usuall y occurre d due t o th e abutment s an d wing walls being pushe d
backwards an d forwards . Althoug h admittin g inadequat e knowledg e abou t th e
extent o f th e movements , an d th e lengt h o f spa n when expansion become s
important, i t wa s recommende d tha t expansio n joint s ar e not generall y neede d
for span s o f u p t o 9m . Fo r reinforced concret e a temperatur e range o f 33° C
was suggeste d a s being reasonabl e an d too k int o accoun t the fac t tha t the ful l
range o f ai r temperatures would not b e experience d b y th e superstructur e due
to th e lo w therma l conductivity o f th e concrete . Thi s would give a movemen t
of 11m m per 30 m length. Contemporar y practice was t o fi t joint s a t th e end s
of bridge s bu t fo r bridge s longe r tha n 42m joints wer e sometime s positione d
over th e piers .
Chettoe an d Adam s (1938 ) outline d th e type s o f expansio n join t an d
bearing the n bein g used: slidin g plates, simpl e cas t iron rockers , heavy cas t
iron o r cast stee l rockers , rollers, reinforced concret e rockers.
4 Recen t development s 195 0 - presen t
The us e o f continuou s superstructure s recommended b y Bresse y was no t

generally adopte d possibl y because o f factor s such a s th e uncertainit y o f
differential settlemen t an d th e greate r economy o f using standard beams with
simply supporte d spans . Afte r th e hal t i n ne w constructio n durin g th e 1940' s
the motorwa y er a starting in th e lat e 1950' s contained example s o f long
viaducts havin g simpl y supporte d spans , fo r exampl e th e Midlan d Links . Ther e

were als o som e notabl e continuou s structures , for exampl e Westwa y an d
Hammersmith.
In th e 1970' s more comprehensiv e measurement s o f bridge temperatures
were mad e fo r differen t type s o f bridge s b y th e Transpor t and Roa d Research
Laboratory an d i t becam e eviden t tha t ranges o f temperatur e are much greater
than hithert o considered . Thi s produce d highe r calculate d values o f expansio n
and furthe r discourage d desig n of lon g continuou s spans .
With th e developmen t o f highe r strengt h concret e and mortar s
there ha s bee n a loss in th e ductilit y whic h enable d earlie r structures to cop e
with thermal effects . Thi s i s well illustrate d b y brick and ston e structures
which unti l comparativel y recent times di d no t requir e expansion joints . I n
contrast desig n of brickwor k now require s 20m m joints a t 10 m intervals
containing a fille r capabl e o f compressin g t o 10mm . Thi s procedur e is eve n
adopted fo r no n structura l brickwork such a s garden walls. Th e efficac y o f
waterproofing th e joints i s dependen t o n th e fille r adherin g t o th e brickwork ; i n
practice thi s i s a commo n sourc e o f failur e an d ther e ar e examples o f splittin g
at th e interfac e afte r very shor t times .
In highway bridge s failure s o f expansio n joint s becam e eviden t a t a n early
stage bu t fo r som e tim e th e resultin g leakag e wa s considere d t o b e onl y a
nuisance permittin g unsightl y stain s t o develop . I n the tim e whe n gritting wa s
used t o comba t th e effect s o f ic e an d sno w there were n o significan t problem s
but thi s change d afte r th e introductio n o f deicin g sal t i n th e mid-1950's . Th e
first sign s o f corrosio n problems wer e i n decks , particularl y in th e Unite d
States, whe n it wa s foun d tha t chloride s fro m deicin g sal t were corrodin g th e
steel reinforcement . A fe w year s later it became eviden t tha t leakage throug h
expansion joint s was , dependin g o n th e desig n of th e bridg e i n question ,
permitting sal t wate r to flo w ove r th e substructur e causing corrosio n to th e
reinforcing stee l i n crosshead s an d piers. Les s commonl y ther e was corrosio n
of anchorage s an d tendon s o f pos t tensione d beams . I n a surve y o f th e
performance o f concret e in 20 0 of th e Departmen t o f Transport' s bridges ,
Wallbank (1989 ) reporte d that three quarter s of al l joints wer e leaking .
The problem s cause d b y leakag e ha s focuse d attentio n o n th e nee d t o
develop mor e durabl e design s o f expansio n joints , particularl y in relatio n t o
their abilit y t o resis t th e actio n o f fatigu e an d remai n watertight. I n thi s
context severa l investigation s ar e now bein g conducted t o determin e fatigu e
performances an d improv e designs .
A ver y differen t approac h is t o desig n continuous bridge s t o avoi d th e nee d
for expansio n joint s ove r th e piers . Ther e ar e two categorie s o f continuity ;
continuous superstructure s having expansio n joint s a t their end s an d integra l
bridges havin g superstructure s that are either fixe d t o thei r abutment s a s i n
portal frame s o r free t o slid e s o tha t the en d o f th e dec k i s effectivel y absorbe d
into th e pavement . Moder n jointless bridges , a s oppose d t o th e earlie r type s
built i n ignorance , hav e bee n constructe d in th e Unite d State s i n th e las t 2 0
years, particularl y in Tennessee . T o dat e th e performanc e o f thes e bridge s ha s
52 Tilly
been sai d t o b e ver y good an d ther e has been n o seriou s crackin g or

degradation reported . It is notabl e tha t in thi s tim e ther e would almos t
certainly hav e bee n failure s o f expansio n joints . A s wit h the earlie r masonry
viaducts, integra l bridges appea r to def y th e law s o f physics . Ther e ar e several
possible mechanism s tha t may operat e to permi t integra l bridges t o work:
• Som e o f th e movemen t ma y be transferre d to a rockin g actio n o n th e

abutments.
• Th e ful l temperatur e range is usuall y experience d ove r a relativel y lon g
period o f tim e an d thi s woul d enabl e th e structur e to creep.
• Th e structur e may be abl e t o cop e wit h the mor e rapi d ranges o f
temperature b y distortion .
In th e lon g term jointless bridge s constructe d with modern less forgivin g

materials ar e liable t o exhibi t distres s throug h fatigu e crackin g of th e concret e
but thi s i s likel y to b e i n time s comparabl e with the desig n life. Thi s i s
exemplified b y recen t works t o refurbis h a 6 0 year old 4-spa n bridg e havin g a
total lengt h o f 65m . Th e masonr y piers an d abutment s wer e fro m a n earlier
bridge constructe d in 1875 . Th e dec k wa s o f continuou s reinforce d beam-and -
slab constructio n an d n o provisio n ha d been mad e fo r expansion . Ove r th e
years th e masonr y abutment s becam e cracke d and deteriorate d to th e exten t
that extensiv e repair s were required . Nevertheless , i t ha d performe d
satisfactorily withou t th e problem s pose d b y expansio n joints .
5 Concludin g remark s
In th e earl y types o f structur e it was possibl e t o ignor e effect s o f expansio n

because constructio n materials were mor e ductil e an d forgiving , Moreover ,
components wer e smalle r s o tha t relative movement s an d multi-hinge s coul d
form enablin g th e structur e to deform . Moder n high strength material s enabl e
longer span s an d lighte r weight s t o be achieve d bu t structure s are less abl e t o
deform.
It was prove d virtuall y impossibl e t o devis e expansio n joint s tha t remain
serviceable an d water tight fo r a reasonabl e perio d o f time . I n consequenc e
the maintenanc e cost s o f moder n bridge s ar e increased du e t o th e cos t o f
repair works an d consequentia l delay s t o traffic .
Experience o f reinforce d concret e bridges constructe d in th e earl y years o f
this centur y has show n that jointless constructio n can last fo r 6 0 years or more.
In case s wher e damag e occurre d it appear s to hav e bee n du e mainl y t o fatigu e
processes. Suc h damag e ma y have bee n les s expensiv e t o repai r than
articulated bridge s havin g faile d expansio n joint s an d consequentia l corrosion .
Experience i n th e Unite d States , ove r th e las t 2 0 years indicates tha t it i s
possible t o operat e long spa n concret e bridges havin g n o expansio n joints ,
without apparen t distress .
It appear s tha t curren t allowances fo r temperatur e effects ma y b e excessiv e

and ther e i s scop e t o reconside r method s o f designin g to cop e wit h
temperature effects .
6 Reference s
Bressey, C.H . (1933 ) Th e Developmen t o f Roa d Bridge s i n Grea t Britain.

British Bridges , p p vii-xxiv , Publi c Works , Roads an d Transport
Congress.
Chettoe, C.S . an d Adam , H.C . (1938 ) Concret e Bridge Design .
Drysdale Dempsey , G . Tubula r Bridges, (1864) . Virtu e Brother s & Co.
Hunter A . (1920 ) Bridg e an d Structura l Engineering Handbook , Si r Willia m
Arrol, Spon .
Page, J . (1987) Loa d tests t o collaps e o n tw o arc h bridges a t Preston,
Shropshire an d Prestwood , Staffordshire . Departmen t o f Transpor t T R R L
Report, RR110 , T R R L Crowthorne .
Shirley Smith , H . (1954 ) Th e World s Grea t Bridges. Phoenix .
Waddell. (1916 ) Bridg e Engineering . Chapma n an d Hall .
Wallbank, E.J . (1989 ) Th e Performanc e o f Concret e Bridges , HMSO .
6 ADVANTAGE S O F
BRIDGE DEC K
CONTINUITY
B.P. PRITCHAR D
Consultant to W.S. Atkins Consultants Ltd and
Colebrand Ltd, Haywards Heath, UK
Abstract
Multispan bridg e deck s whic h ar e designed to provid e continuit y acros s the suppor t
piers offe r a number o f advantages , bot h i n th e long-ter m performanc e o f th e
structure, bu t als o i n simplificatio n o f th e severa l aspect s o f construction , an d
reduced overal l constructio n costs . Thi s pape r outline s thes e advantage s an d gives
examples o f comparativ e simpl y supporte d an d continuou s designs .
Keywords: Dec k continuity ; reductio n o f joints; reductio n i n dept h o f flyover s an d
underpasses; reductio n i n lengt h an d heigh t o f flyover an d approac h ramps;
reduction o f bearings ; constructio n depth ; pie r an d foundatio n savings ; securit y o f
extra redundancy .
This pape r has bee n prepare d from Chapte r 2 o f th e author' s book 'Bridg e Design fo r Econom y &
Durability', publishe d i n 1992 . Thank s are due t o th e publishers , Thoma s Telford Publications , for
permission t o reproduc e it. Th e boo k i s availabl e from th e Thoma s Telford Bookshop , I C E,
London o r fro m Thoma s Telford Service s Ltd, Thomas Telford House, 1 Heron Quay , London E 14
4JD.
1 Dec k continuit y
Early multi-spa n bea m an d sla b o r slab bridge s wer e generall y designe d an d buil t
as a sequenc e o f simpl y supporte d spans . I n 193 0 the American , Hardy Cross,
introduced hi s famou s momen t distributio n metho d fo r th e quic k an d eas y analysi s
of continuou s frame s an d beams . Thi s metho d wa s immediatel y take n o n b y
1
bridge designer s an d a gradua l build-up i n th e us e o f dec k continuit y commenced .

The 1950 s and 1960 s saw a further great improvement i n continuous-dec k
analysis wit h the introductio n o f computers , togethe r with grillage an d finit e
element programs . Th e us e o f continuit y spread , particularl y in winter-affecte d
countries wher e th e maintenanc e problem s o f dec k joints an d de-icin g sal t hav e
dramatically multiplied . I n som e countrie s nationa l bridg e code s recommen d o r
even dictat e multi-spa n dec k continuit y unles s ground settlemen t problem s requir e
the greate r articulation provided b y a serie s o f simpl y supporte d spans .
Currently, th e U K Departmen t o f Transpor t is considerin g th e issu e o f a n
Advice Not e encouragin g bridg e designer s t o us e dec k continuity , principall y t o

56 Pritchard
Fig. 1 A 5 flyover , Staple s Corne r Interchange , London.

The joint-free dec k act s as a n umbrella ove r th e pier .
minimise th e us e o f trouble-pron e dec k joints. Th e not e wil l also recommen d

'integral bridges', i n which th e continuit y o f bridge deck s fre e o f intermediat e joint s
is extende d t o th e eliminatio n o f th e en d expansio n joint s b y building the dec k int o
the abutments. 2
The eliminatio n o f dec k joints i s a good an d sufficien t reaso n fo r adoptin g

deck continuity , wit h al l th e implie d saving s i n maintenanc e costs . However , i t
should b e remembere d tha t dec k continuit y ca n als o offe r considerabl e economie s
in first-tim e constructio n costs . I t is thi s aspec t which i s emphasise d i n section s 2
and 3.
Continuity ca n be adopte d fo r mos t multi-spa n bridges . Onl y i n very short
span structure s and i n minin g area s will differential pie r an d abutmen t settlement s
become significan t enough t o dictat e a chang e t o simpl y supporte d spans .
Differential settlement s i n highwa y bridge s rarel y excee d 10-2 0 mm an d i n
the mor e critica l shorter spans ther e is a tendenc y t o us e reinforce d concret e deck s
anyway. Becaus e o f th e lowe r concret e elasticity modulu s an d inherent cracking ,
reinforced concret e is less stif f tha n prestresse d concret e and accommodate s
differential settlemen t wit h less attracte d moment .
In certai n cases single-spa n o r three-span deck s subjec t t o significan t
settlements ca n be give n some o f th e benefit s o f continuit y b y counterweighting . 3
Advances of bridge deck continuity 5 7
2 Advantage s
Figures 1 & 2 illustrat e th e mai n advantage s t o be gaine d b y using multi-span dec k

continuity: Fig . 2(a ) shows a typica l simpl y supporte d multi-spa n deck , Fig . 2(b ) a
corresponding continuou s deck .
The firs t advantag e is th e eliminatio n o f dec k an d parape t joints a t th e
piers. Thi s represent s saving s i n bot h constructio n an d maintenance cost s -
maintenance whic h include s tha t o f th e underlyin g piers . Eliminatio n o f joints als o
gives th e vehicle s crossin g th e bridg e a better ride.
The secon d advantag e i s th e savin g i n dec k depth , whic h no t onl y ca n
reduce th e constructio n cos t o f bot h th e dec k an d th e supportin g substructur e but
also ca n reduce th e lengt h o f th e dec k o r the associate d approac h ramps or
earthworks. Th e latte r savin g i s probabl y th e mos t significan t construction-cos t
economy an d i s deal t wit h separately in sectio n 3 below.
The thir d advantag e lies in th e halvin g o f th e numbe r o f bearings require d
at eac h pier . Althoug h th e bearing s will be o f highe r capacity , the lesse r numbe r
should resul t i n cos t savings .
The fourt h advantag e result s fro m th e replacemen t o f th e tw o row s o f
bearings b y th e singl e ro w o f centra l bearings. Th e pier s o r crossheads ca n b e
Road leve l Joints
(a)
d<
2 cfi =
b<2 £>i
(b)
Fig. 2 Advantage s o f continuity : (a ) simply supporte d deck ; (b ) continuou s deck .

58 Pritchard
reduced i n thicknes s no t onl y becaus e th e singl e row o f bearings take s u p less roo m
at th e top , but becaus e th e dec k dea d loa d ha s reduce d an d th e liv e load moment s
applied b y off-centr e pair s of bearing s ar e removed. Significan t cos t saving s i n pie r
foundations als o result .
A fift h advantage , possible onl y wit h deck continuity , i s tha t the 'sucker
deck' principl e ca n be applie d t o advantage . 3
A sixt h advantage , agai n possibl e onl y with deck continuity , i s tha t use ca n
be mad e o f th e transvers e strength o f th e dec k ove r pier s (equa l t o th e longitudina l
strength for sla b decks ) t o provid e inboar d piers an d singl e bearing support .
Considerable benefit s ca n result fro m simplificatio n o f ske w crossings an d lan d
savings b y passin g interchang e sli p roads under th e mai n flyover . 4
A sevent h advantag e i s tha t continuit y mean s extr a dec k redundancy , whic h

generally guard s agains t sudde n collaps e i n overloa d situation s o r in instance s o f
sabotage (Fig . 3).
Fig. 3 Bom b damag e t o A 5 flyover , Staple s Corne r Interchange.

3 Saving s i n dec k constructio n dept h
The optimu m constructio n depth/spa n ratio s vary with span, liv e loading, material s
and method s o f construction . Averag e ratio s for th e 10-6 0 m spa n rang e are
about 1/2 4 for continuou s span s an d 1/1 8 for simpl y supporte d spans . Thus , fo r
35-40 m span s abou t a half metr e ca n be save d i n dec k constructio n dept h b y
adopting continuity . I n itself, th e reductio n i n concret e or steel woul d appea r to
lead t o a reductio n i n constructio n cost . Thi s i s no t alway s th e case , a s continuit y
can introduc e extr a complication s t o th e sequenc e an d standardisatio n o f
construction, usuall y mor e perceive d tha n real if desig n and constructio n are
adequately planne d t o sui t th e continuity .
Whether th e continuou s dec k i s cheape r to build or not, th e associate d
saving i n constructio n dept h ca n certainly reduce constructio n cost s o f othe r
elements o f th e bridg e o r viaduct. Th e substructure , of course , ca n benefit b y
reductions o f u p t o 10-15 % i n dec k weight, significan t wit h the longe r span s an d
where foundation s ar e poor. However , i t i s th e possibl e reductions i n relate d
earthworks, approac h ramps or bridge length s whic h ar e most cost-effective .
Typical example s ar e demonstrated in Fig s 4-6.
Figure 4 show s a typica l urban flyover carryin g a highway ove r a ground -
level roundabout . Th e approac h ramps rise a t a 4 % grade and th e 18 0 m long
viaduct ha s a complian t vertica l curve providing th e require d headroom ove r th e
underlying roundabou t carriageways . I n Fig. 4(a ) there ar e five equa l 3 6 m simpl y
supported span s wit h a dec k constructio n depth o f 2 m. Th e resultin g maximu m
ramp heigh t i s 7 m , givin g a n overal l lengt h o f 17 5 m fo r eac h ramp .
Roundabout
carriageway
(b)
Fig. 4 Approac h ramp savings: (a ) simply supporte d spans ; (b ) continuou s span s

60 Pritchard
Fig. 5 M l flyover , Hendo n Fiveway s Interchange , London.
The proposa l i n Fig , 4(b ) uses spa n continuit y an d slightl y modifies th e spa n
layouts t o giv e th e mor e balanced arrangemen t of shorte r end spans . Thi s result s
in thre e inne r span s o f 3 8 m an d tw o en d span s o f 3 3 m. Th e associate d reductio n
in dec k dept h o f 0. 5 m mean s tha t the maximu m ram p height drop s t o 6. 5 m an d
the ram p length reduce s t o 162. 5 m.
The tota l reductio n i n th e approac h ramps due t o th e introductio n o f
viaduct continuit y i s therefor e measure d by a 7 % reduction i n lengt h an d height .
This result s i n a reductio n o f nearl y 14 % in ram p wall area , whic h als o give s a
rough approximatio n o f th e cos t savin g i n ram p construction a t eac h end . Fig . 5
shows a n earl y flyover ove r th e M l whic h benefited wit h significant ram p saving s
generated b y th e dept h reductio n i n th e continuou s decks .
Figure 6 show s another flyover wit h a simila r vertical profile, thi s tim e
crossing a centrall y located urba n carriageway flanked b y pedestria n areas . I n thi s
case, th e designe r agai n choose s t o us e approac h ramps rather tha n embankment s
to reduc e landtake . Hi s calculation s sho w that the ramp s can rise t o a height o f
8 m usin g sprea d footings fo r th e ram p walls, whil e remainin g cheape r per square
metre tha n suspende d deck . Anythin g highe r would require expensiv e piling ,
exceeding th e cos t o f suspende d deck . Th e maximu m lengt h o f eac h o f th e 4%
ramps i s therefor e se t a t 20 0 m.
A 3 6 m spa n i s require d acros s the centra l carriageway and Fig. 6(a ) shows
a sequenc e o f thre e equa l 3 6 m simpl y supporte d spans , th e oute r span s crossin g
Approach
Abnormal
clearance
83 m
(b)
Fig. 6 Saving s i n flyover length : (a ) simpl y supporte d spans ; (b ) continuou s spans .
the pave d pedestria n areas . Th e 2 m dec k constructio n dept h result s i n th e

maximum allowabl e ram p height o f 8 m an d length o f 20 0 m.
Fig. 6(b ) shows the effec t o f providin g spa n continuit y an d th e associate d
0.5 m savin g i n dec k constructio n depth . Th e 20 0 m long, 8 m hig h ramps are
retained, bu t eac h ca n now b e relocate d 12. 5 m neare r the centra l carriageway,
reducing th e require d viaduct lengt h fro m 10 8 m t o 8 3 m an d th e en d span s fro m
36 m t o 23. 5 m. Thus , with a smal l los s o f pedestria n are a an d no chang e i n th e
approach ramps , the flyover viaduc t lengt h i s reduce d by som e 24% , with simila r
percentage saving s i n viaduc t cos t and , possibly , landtake . Fo r a dua l three-lan e
motorway flyover, curren t deck cost s would indicate a savin g o f som e
£400,000-£500,000.
The fina l example , show n in Fig . 7 , i s a dua l three-lan e motorwa y locall y
depressed t o pas s unde r twi n 36 m spa n simpl y supporte d bridge s carryin g a 16 0 m
diameter groun d leve l roundabout . I n thi s cas e th e vertical profile o f th e motorwa y
consists o f a n 8 0 m lon g cres t curve followed by a 4 % downgrade sectio n 15 0 m
long, the n a sa g curv e 16 0 m long followed by 15 0 m o f 4 % upgrade an d a fina l
80 m lon g cres t curve . Th e bridge s ar e located just abov e th e end s o f th e sa g
curve, whic h ar e 7.3 m below ground level . Thi s dept h provide s th e require d 5. 3 m
clearance unde r th e 2 m dec k constructio n depth .
The depresse d sectio n o f motorwa y is 62 0 m long an d th e excavatio n
required with 1 : 2 cuttin g slope s will be aroun d 150,00 0 m . Saving s ca n be mad e
3
by addin g continuit y t o th e roundabou t bridg e decks . Th e 0. 5 m dec k constructio n

depth save d allow s the depresse d motorwa y profile t o be raise d by th e sam e
amount, show n a s a broken lin e o n th e figure . Thi s save s approximatel y 15,00 0 m , 3
62 Pritchard
Bridge Roundabou t
620 m lon g depressed motorway
Fig. 7 Excavatio n savings, depresse d motorway .
or 10% , of th e excavation . Ther e ar e also smal l saving s i n lan d acquisitio n and ,

indeed, i n th e overal l lengt h o f th e roundabou t overbridges .
Depending o n th e excavate d material and its eas e o f disposal , th e excavatio n
cost saving s whic h resul t fro m changin g th e roundabou t bridg e dec k fro m simpl y
supported t o continuou s ca n readily excee d £100,000 . I n rock, of course , th e savin g
could b e man y time s higher . Fig . 8 show s suc h a depresse d interchange , i n thi s
case with a n additiona l top-laye r flyover .
Fig. 8 Depresse d M2 5 Interchange.

1 Reference s
1 Cross , H. Analysi s o f continuou s frame s by distributin g fixed-en d moments .

Proc. Am. Soc. Civ. Engrs, 1930.
2 Burke , M. Th e integrate d constructio n an d conversio n o f singl e an d
multispan bridges . Bridge Management, Proceeding s o f International
Symposium, Surre y University, 1990 . Elsevier .
3 Pritchard , B . P., Haunched bridge s an d th e Sucke r Deck Principle ,
Proceedings of the Institute of Civil Engineers, Structures and Buildings,
February 1994.
4 Pritchard , B . P. Bridge design for economy and durability. Thoma s Telford ,
London, 1991.
7 C O N T I N U I T Y I N DECK S
WITH PRECAS T BEAM S
- PRACTICA L ISSUE S
H.P.J. T A Y L O R
Costain Building Products Ltd, Stamford, UK
Abstract
T h i s p a p e r c o n s i d e r s t h e d e s i g n i s s u e s i n v o l v e d i n makin g f u l l y
c o n t i n u o u s d e c k s f r o m s i m p l y s u p p o r t e d b r i d g e bea m a r r a n g e m e n t s an d
the p r a c t i c a l c o n s i d e r a t i o n s t h a t the y impose .
Keywords C o n t i n u i t y , design , limi t s t a t e s constructio n joints .
1 Introductio n
F a c t o r y mad e p r e t e n s i o n e d p r e s t r e s s e d c o n c r e t e b r i d g e beam s h a v e
b e e n on e o f t h e s u c c e s s s t o r i e s i n b r i d g e s o v e r t h e l a s t 4 0 y e a r s .
P a r t i c u l a r l y i n t h e U K a n d t h e US A man y t h o u s a n d b r i d g e d e c k s u s i n g
t h e s e b e a m s , a s i n f i l s l a b d e c k s , bea m a n d s l a b d e c k s a n d v o i d e d
d e c k s , h a v e b e e n c o n s t r u c t e d w i t h r e m a r k a b l y fe w p e r f o r m a n c e
d i f f i c u l t i e s w i t h r e s p e c t t o t h e beam s a n d d e c k s . Th e mai n a r e a o f
l a c k o f goo d d u r a b i l i t y h a s bee n i n t h e weaknes s o f t h e v a r i o u s form s
o f j o i n t s i n w h a t a r e u s u a l l y s i m p l y s u p p o r t e d d e c k s . Th e obviou s
a n s w e r t o t h i s i s c o n t i n u i t y o f som e f o r m .
C o n t i n u i t y i n b r i d g e d e c k s w i t h p r e t e n s i o n e d p r e c a s t beam s i s n o t
new, a s i g n i f i c a n t a m o u n t o f r e s e a r c h a n d d e s i g n d e v e l o p m e n t d a t a
e x i s t s i n th e l i t e r a t u r e , datin g bac k t o th e e a r l y 1 9 6 0 ' s .
2 P r a c t i c a l Issue s
F r o m t h e p o i n t o f v i e w o f e a s e o f c o n s t r u c t u c t i o n an d economy , t h e
most s u i t a b l e metho d o f c o n s t r u c t i n g a c o n t i n u o u s j o i n t i s t o
m a x i m i s e t h e p r e c a s t e l e m e n t an d m i n i m i s e t h e s i t e wor k an d t h e
number o f b e a r i n g s , F i g . 1 ( a ) .
The n a r r o w i n s i t u c o n n e c t i o n o v e r a s i n g l e l i n e o f bearing s
produces th e followin g advantages .
• B e a r i n g number s a r e m i n i m i s e d , b e a r i n g s a r e e x p e n s i v e
• Beam s a n d i n s i t u s h u t t e r c a n a l l b e s u p p o r t e d f r o m t h e m a i n p i l e
cap
I n s i t u c o n c r e t e i s minimise d
S h u t t e r i n g i s minimise d
The m e t h o d i s a p p r o p r i a t e f o r w a l l o r c o l u m n s u p p o r t
Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published i n 1 9 9 4

by Taylo r & Francis, 2 Park Square, Milton Park , Abingdon, Oxon , OX14 4RN. ISBN: 0 419 19030 9 .
66 Taylor
T h i s f o r m o f c o n s t r u c t i o n r e q u i r e s t h a t tw o f e a t u r e s a r e p r e s e n t .
The s p a n s s h o u l d b e s u c h t h a t t h e p r e c a s t beam s a r e w i t h i n t h e
t r a n s p o r t a b l e r a n g e an d t h e d e s i g n s h o u l d b e s u c h t h a t t h e r e i s n o t
t o o muc h c o n g e s t i o n o f r e i n f o r c e m e n t i n t h e i n s i t u j o i n t . Th e us e o f
two l i n e s o f b e a r i n g s s e e m s t o b e u n n e c e s s a r y F i g . 1 ( b )
I n t h i s c a s e , movemen t o f t h e d e c k a s t h e l o a d e d l e n g t h c h a n g e s
f r o m o n e s p a n t o a n o t h e r ma y r e q u i r e t h a t e a c h b e a r i n g b e a b l e t o
c a r r y m o r e t h a n h a l f t h e l o a d f o r a s i n g l e l i n e , som e d e s i g n e r s mak e
each l i n e capabl e o f carryin g th e f u l l loa d wit h a consequentia l
increase o f cost .
W i d e r j o i n t s ma y b e u s e d a n d ma y b e a p p r o p r i a t e w i t h ske w d e c k s o r
with a bridg e curvin g i n plan . I n t h i s cas e i t i s importan t t o
c o n s i d e r t h e r e l a t i v e s t i f f n e s s o f t h e s u p p o r t o f t h e beam s a n d t h e
insitu concret e shutter . Settlemen t durin g th e castin g operatio n
c o u l d r e s u l t i n p l a s t i c c r a c k s i n t h e i n s i t u c o n c r e t e an d i n e x t r e m e
c a s e s , c r a c k s a t t h e t o p o f t h e p r e c a s t beams . S h u t t e r suppor t i n
t h e f o r m show n i n F i g . 2 ( b ) s h o u l d b e a v o i d e d , t h e i d e a l m e t h o d i s t o
u s e t h e f o r m show n i n F i g . 2 ( a ) .
/ / / / / // / /
Fig. 1. Minimizing the number of bearings

Continuity in decks with precast beams 6 7
Fig. 2. Shutter support for wider beams.
3 Desig n Consideration s
The i n t e g r a l j o i n t o v e r a s u p p o r t c o n s i s t s o f p r e c a s t p r e s t r e s s e d
beams w i t h a r e i n f o r c e d c o n c r e t e s l a b a n d i n f i l d i a p h r a g m . T h i s
c r e a t e s a h y b r i d s e c t i o n w h i c h ma y p r e s e n t som e c o n f u s i o n i n
analysis.
SLS H o g g i n g B e n d i n g
At t h e s e r v i c e a b i l i t y l i m i t s t a t e (SLS ) t h e s e c t i o n ma y b e
s u b j e c t e d t o h o g g i n g moment s F i g . 3 . I n thi s c a s e , i n BS540 0 th e to p
slab i s considere d t o b e reinforced , wit h a c r a c k widt h c r i t e r i o n
a n d t h e p r e s t r e s s e d bea m p r e s t r e s s e d w i t h a n o tensio n c r i t e r i o n .
The d e s i g n e r h a s t o d e c i d e w h e t h e r c r a c k s i n t h e s l a b ma y b e
68 Taylor
Fig. 3 . Hogging at SLS and ULS.
p e r m i t t e d t o p r o g r e s s dow n i n t o t h e t o p o f t h e p r e s t r e s s e d bea m a t
t h e SLS . Suc h c r a c k i n g c a n n o t b e i m p o r t a n t fro m t h e d u r a b i l i t y p o i n t
o f v i e w a s i t woul d n o t p a s s t h r o u g h an y o f t h e p r e s t r e s s e d s t r a n d s .
R e s e a r c h o n c o m p o s i t e c o n s t r u c t i o n h a s show n t h a t a t a j o i n t b e t w e e n
p r e c a s t an d i n s i t u c o n c r e t e t h e t e n s i l e c r a c k i n g s t r a i n an d
c r a c k i n g b e h a v i o r o f th e weake r s e c t i o n i s c o n t r o l l e d b y th e
s t r o n g e r . Desig n s t u d i e s an d a revie w o f th e l i t e r a t u r e o n
composite c o n s t r u c t i o n ar e require d i n t h i s are a i n orde r tha t
p o s i t i v e a d v i c e ma y b e g i v e n t o d e s i g n e r s , w i t h i n t h e r e q u i r e m e n t s o f
BS5400.
ULS H o g g i n g B e n d i n g
A f u r t h e r c o n s i d e r a t i o n o f t h e sam e s e c t i o n i s h o g g i n g b e n d i n g a t
t h e ULS . B S 5 4 0 0 c o n t a i n s method s f o r t h e c a l c u l a t i o n o f t h e u l t i m a t e
s t r e n g t h o f s e c t i o n s whic h c a n t a k e a c c o u n t o f s t r e s s e d an d
u n s t r e s s e d r e i n f o r c e m e n t i n b o t h t h e t e n s i l e an d c o m p r e s s i v e z o n e .
I t i s p o s s i b l e t o modif y t h e c o n c r e t e s t r e s s b l o c k t o t a k e a c c o u n t o f
t h e p r e c o m p r e s s i o n o n t h e b o t t o m f l a n g e o f a p r e s t r e s s e d bea m b u t
d e s i g n e r s a r e u n f a m i l i a r w i t h t h e c o n c e p t . Th e strai n compatibilit y
m e t h o d , i l l u s t r a t e d i n F i g . 4 i s w e l l know n a n d t h e m o d i f i c a t i o n s
shown i n t h e f i g u r e t o m o d i f y t h e s t r e s s b l o c k s h o u l d b e d e v e l o p e d
a n d c o m p a r e d w i t h t h e l i t e r a t u r e . Th e m e t h o d , whe n u s e d t o c a l c u l a t e
t h e s t r e n g t h o f t h e s e c t i o n t e s t e d b y R e g a n , show s g o o d f i t .
1
Parasitic Effect s
The p r e s t r e s s e d b e a m s w i l l c o n t i n u e t o d e f o r m du e t o l o n g t e r m
c r e e p a n d d i f f e r e n t i a l s h r i n k a g e a f t e r t h e d e c k i s mad e c o n t i n u o u s .
T h i s e f f e c t w i l l p r o d u c e p a r a s i t i c m o m e n t s , a s show n i n F i g . 5 a n d i n
t h e s e t o f b e n d i n g momen t d i a g r a m s i n F i g , 6 . T h i s figur e give s a
summary o f t h e moment s p r o d u c e d i n a n a c t u a l c a s e o f a t h r e e s p a n
b r i d g e a n d show s t h a t t h e e f f e c t i s s i g n i f i c a n t a n d c a n g o a s f a r a s
p r o d u c i n g n e t s a g g i n g moment s a t t h e s u p p o r t i n t h e u n l o a d e d c a s e .
T h i s e f f e c t h a s a l s o b e e n s t u d i e d i n t h e US A i n t h e e a r l y 1 9 6 0 ' s . 2
The r e i n f o r c e m e n t d e t a i l s f o r t h i s c a n b e d i f f i c u l t t o c o n t a i n w i t h i n
a narro w j o i n t .
An a l t e r n a t i v e t o t h i s i s n o t t o a t t e m p t t o r e i n f o r c e b u t t o a l l o w
t h e j o i n t t o c r a c k i n t h i s s i t u a t i o n , r e l i e v i n g t h e p a r a s i t i c momen t
but p r o d u c i n g a c r a c k e x t e n d i n g fro m t h e s u p p o r t , upward s t o w a r d s t h e
t o p o f t h e b r i d g e . Suc
3
h a c r a c k doe s no t expos e c r i t i c a l element s
prestrain
— H 0.003 5
prestra in
stress-strain stres s bloc k
steel concret e
Fig. 4. Strain compatibility method accounting for prestrained

concrete and steel.
70 Taylor
Fig. 5 . Possible sagging at SLS.
o f t h e j o i n t t o i n g r e s s o f w a t e r a n d s a l t f r o m a b o v e an d t h e r e i s a
l a r g e b o d y o f e v i d e n c e t o sho w t h a t i t d o e s n o t p r o d u c e a d u r a b i l i t y
problems.
T h i s s u b j e c t s h o u l d b e r e v i e w e d an d r e s e a r c h e d t o p r o d u c e soun d
a d v i c e w i t h i n t h e p e r m i t t e d e x t e n t o f B S 5 4 0 0 an d s h o u l d b e d e v e l o p e d
t o a simpl e desig n p r o c e s s t h a t ca n e a s i l y b e f i t t e d i n t o c u r r e n t
c o m p u t e r s y s t e m s f o r g r i l l a g e an d f i n i t e e l e m e n t a n a l y s i s .
References
Regan, P . E . B e h a v i o u r o f P r e c a s t P r e s t r e s s e d Y-beam s i n S h e a r ,
T o r s i o n an d N e g a t i v e Bendin g S t r u c t u r a l E n g i n e e r Vo l 6 8 N o 2 3
December 1 9 9 0
K a r r , P . H . , K r i z , L . B . an d H o g n e s t a d , E . P r e c a s t P r e s t r e s s e d C o n c r e t e
B r i d g e s P i l o t T e s t s o n C o n t i n u o u s G i r d e r s D e v e l o p m e n t B u l l e t i n D3 4
PCA I l l i n i o i s Ma y 1 9 6 0
Hambly, E . an d N i c h o l s o n , B . P r e c a s t I n t e g r a l B r i d g e s P r e s t r e s s e d
Concrete Associatio n L e i c e s t e r 199 0
Initial no n continuous cond 11ion I
Spans
f 22 m f t 28 m H 22 m f
Prestress H h H h H h
Roughly balance d b y dead loa d
Dsad loa d
[Make continuous (sa y at 50 days) 560
Dead load , dec k concret e on

continuous structur e and
with prop s remove d
Superimposed dea d loa d
TotaI of above
Long ter m creep , afte r 5 0 days

Joint A
Long ter m creep / afte r 5 0 days

Joint B
TotaI of above
Indicative moments are shown in

kNm/m of width
Fig. 6. Stages in design of a typical precast, prestressed beam

bridge, made continuous.
R E P O R T O N SESSIO N TW O
S. S H A N M U G AM
Dr Till y presente d a n interestin g accoun t o f th e histor y o f bridg e continuit y an d

showed severa l example s o f earl y bridges datin g from 160 0 BC built i n stone , brick
masonry, iron , concrete and steel t o th e more modern design s pos t 1950 . H e wen t
on t o explai n th e possibl e mechanism s outline d i n hi s pape r togethe r wit h th e
reasons why suc h earl y bridges were abl e to cop e with th e temperatur e effects . H e
considered that there was a need fo r a review o f the current allowance s being mad e
for temperatur e effects i n th e light o f satisfactory performance of so man y jointless
bridges i n recen t years.
In introducing his paper , Mr Pritchard showed severa l stark examples o f bad
details suc h a s leakin g joints , corrode d bearings, an d deterioratio n o f pie r cros s
heads, arisin g fro m th e lac k o f bridg e continuity . H e highlighte d man y inheren t
advantages of multi-span dec k continuity by showing actua l continuous spa n layouts
which ha d resulted in large savings in both constructio n and maintenance costs. H e
pointed out that the extra deck redundancy in continuous structures would als o guard
against sudden collaps e in the event of sabotage, citing an example where bomb blast
had lifte d th e dec k o f a flyover b y 20 0 mm an d blown hole s i n th e box flanges .
Dr Taylor explained the issues involved i n making decks fully continuous from
simply supporte d bridge beam arrangements . H e describe d researc h programme s
carried out notably by Karr, Kriz and Hognestad on pilot tests on continuous girders,
and by Regan on the behaviour of precast prestressed Y beams in shear, torsion and
negative bending .
Dr Taylo r illustrate d severa l example s o f diaphrag m layout wit h th e mos t
beneficial way s o f supportin g th e precas t beams an d the i n sit u diaphrag m shutter ;
he explaine d ho w th e strength of the reinforced support section in negative bendin g
at Ultimate Limit State, sagging moment at the support under long term creep of the
beams, an d hogging momen t ove r the support at Serviceability Limit State could be
adequately deal t wit h i n practice .
PART THRE E
CONTINUOUS
BRIDGES
8 DESIG N CONCEP T O F TH E
TRANS-TOKYO BA Y
BRIDGE
Y. SHIO I
Design and Engineering Department, Trans-Tokyo
Bay Highway Corporation , Japan
S. NAKAMUR A
Nippon Steel Corporation, Kobe, Japan
(formerly Trans-Tokyo Bay Highway Corporation )
Abstract
This pape r intruduces th e desig n concep t o f th e Trans-Toky o Ba y Bridge , a
group o f continuou s girde r bridges. Structura l forms o f th e super -
structures, sub-structures , bearing s an d expansion joint s ar e also presented .
Keywords: Continou s girde r bridges, Rubbe r bearings
1. Outlin e o f th e Trans-Toky o Ba y Bridge
The Trans-Toky o Ba y Highwa y (TTBH ) is a 15.1k m long toll highwa y

connecting Kawasak i an d Kisarazu which ar e situated a t both side s o f th e
Tokyo Bay . Th e highwa y consist s o f tw o 10k m long tunnels unde r the
Kawasaki water s wher e marin e traffic i s heavy , an d a 4.5km lon g bridge
over the Kisaraz u waters wher e marin e traffic i s relativel y sparse , a s shown
in fig.l . Tw o man-mad e island s ar e also situate d a t the middl e o f the
tunnel an d at the transitio n poin t o f th e tunne l an d the bridge.
It shoul d be note d that , althoug h mos t o f th e publi c highway s i n Japan
have bee n constructed b y publi c corporations , a semi-public compan y wa s
founded fo r constructio n o f th e TTBH in orde r to utilize financia l an d
human resource s o f bot h publi c an d private bodies .
The highwa y carrie s dual tw o lan e carriagewa y in th e firs t stage , an d
other tw o lane s ar e planned t o b e extende d i n th e futur e dependin g on the
future traffi c flow . Traffi c volum e i s estimate d approximatel y 33,00 0
veh/day whic h i s expecte d t o reac h 64,000 veh/da y i n 2 0 year s after
opening. Desig n vehicle spee d i s 80km/h .
The bridg e i s 4.4k m lon g with 42 piers. Th e navigatio n are a exist s a t
water dept h abou t 20m , where th e spa n i s expande d u p to 240m . Th e
layout o f th e bridg e i s show n in fig. 2 an d fig.l, i n which th e soil laye r i s
also illustrated . Alluvia l san d an d clay layer s (A ) exis t 5 m to 20 m dee p
under the seabed . I n some alluvia l layer s sand s ar e soft wit h N-value o f
the standar d penetration test s unde r 1 0 and the grain s ar e uniform an d
medium size , whic h migh t creat e a liquefaction problem . Belo w th e
alluvial soil s follo w th e diluvia l soi l layer s with different ages , D l t o D4 ,
where sand s an d clays exis t alternately . D 5 laye r observed overal l th e
Continuous and Integral Bridges. Edite d by B.P. Pritchard . Publishe d in 1 9 9 4

Fig.l Trans-Tokyo Bay Highway
F i g . 2 Layout of the TTB Bridge

Design of the Trans-Tokyo Bay Bridge 77
Hw = 3 0 0 0 ^ l 0 5 0 0 f T i m
u u u u u u u u ' U U U LTU U U U U U U U Ur
! !
5 040 6 920 5 040 | |
_ 3 500 j 22 900 ! 3 500

29 900
Fig.3 Steel Girder
highway alignmen t i s considerabl y stif f wit h N-value ove r 50 an d

earthquake shea r wave velocit y ove r 400m/s, and has been take n a s the
input bas e o f seismi c desig n wave .
2. Structura l forms o f th e bridg e
The stee l bo x girder s with orthotopic deck s hav e bee n adopte d a s super -
structure becaus e o f it s ligh t dea d weight an d rapid construction. Th e
maximum we b heigh t i s 10.5m , and the dec k width i s 22.9 m whic h
accommodate dua l tw o lan e carriagewa y in the firs t stage , an d will b e
extended t o allo w th e extr a tw o lane s i n th e future.
Composite stee l pier s wit h steel pile s shown in fig. 4 ar e constructed for
the pier s P I t o PI 2 i n th e 15 m to 25 m deep wate r area an d will suffe r
strong horizonta l force s du e t o ocea n waves an d earthquakes. Stee l pil e
well foundatio n an d concrete column wit h 6 m by 4 m cross sectio n show n in
fig.5 wa s selecte d fo r the pier s P1 3 to P4 2 in th e shallo w wate r areas.
The bridg e i s designe d t o hav e lon g continuous girder s s o a s to increase
the seismi c resistanc e b y sharin g seismi c horizonta l force s wit h many piers,
and to improv e roa d surface smoothnes s b y eliminatin g expansio n joints .
There ar e mainly thre e means t o achiev e continuou s girder s a s shown in
fig.6 (a ) to (d) . I n system (a ) only on e pie r i s fixe d t o the girder , and
therefore horizonta l force s concentrat e on it . Whe n piers ar e flexible ,
system (b ) i s reasonabl e an d more than one pier s ar e fixed an d horizontal
forces ca n be distributed . I n system (d ) dampers d o not resis t slo w motion s
due t o temperatur e changes, wherea s sho w larg e resistance t o quic k motion s
due t o earthquakes . Syste m (c ) utilizes shea r elongations o f rubbe r bearings
to absor b lateral girder motions. Fo r the girder s i n dee p water s where th e
piers ar e high and flexible syste m (b ) i s adopted , an d system (c ) is chose n
for thos e i n shallo w water s where th e pier s ar e relatively lo w an d rigid.
78 Shioi an d Nakam ura
RC S l a b
Base C o n c r e t e
Gravel
Fig. 4 Steel Composite Pier Fig.5 Concrete Pier
(a) (b) (c) (d)
MM F M M M
FF M ! •F F F V M M F M M
tt T T t f f t f f V V H V ij J tg l f U tj
nnnnnnnnnn nnnnnnnnn n
Fig.6 M e a s u r e s to Achieve Multi-Continuous Bridges
Design of the Trans-Tokyo Bay Bridge 7 9
3. Continuou s bridg e wit h flexible pier s
For th e bridg e P 3 to P1 3 the girde r and the piers ar e fixed wit h hing e
joints excep t en d piers . Fig. 7 shows bendin g moment , axia l compressiv e
forces an d shear forces o f girder s du e t o dea d loads , liv e load s an d
temperature chang e o f 35°C . Th e bendin g momen t cause d b y the
temperature chang e i s onl y 5 % o f th e dea d an d live loads . Th e maximum
compressive axia l forc e a t the cente r of th e P7-P 8 du e t o the temperature
change i s abou t 5,500t f whic h i s equivalen t t o 10 % o f th e total stress . Th e
maximum girde r vertical deflectio n du e t o live load s i s 407mm , which i s
within the allowabl e value . Th e maximum vertica l reactio n of bearing s
occurs a t P7 with 3,890tf, an d no lif t forc e exists . Thes e calculatio n result s
above sho w th e validit y i n achievin g continuou s girders .
Fig.8 show s th e dimensio n o f column s an d piles o f P 4 to P12 . Th e

number o f pile s require d fo r th e middl e pier s suc h a s P 7 depends o n the
horizontal force s du e t o earthquakes , whereas thos e fo r en d pier s suc h a s P3
and P10 depend o n the latera l force cause d b y temperatur e change.
4. Continuou s bridg e wit h rubber bearings
For th e bridg e P1 3 to A2 i n shallo w wate r area piers ar e concrete and

relatively stiff , therefor e rubbe r bearings ar e selected t o achiev e multi-spa n
continuous bridge . Thre e types o f rubbe r bearings show n in fig. 9 wer e
thought t o b e feasibl e t o absor b the larg e vertical and horizontal force s a s
well a s lon g displacements : namel y (a ) lead rubbe r bearing (LRB) , (b ) high
damping rubbe r bearing (HDRB) , (c) ring rubbe r bearing(RRB) .
Dynamic analysi s wa s carrie d out t o investigat e th e earthquak e response
to the bridge . Fig.1 0 is a simplified mode l an d fig.ll i s th e design
acceleration wave s adopte d fo r th e TT B structures. Thi s desig n wave i s
input fro m th e bas e stratu m and amplified throug h the soil layers . On e of
the interestin g calculatio n result s i s show n in fig . 12 . Th e peak response i s
0.48Hz a t the girde r center, and is equivalen t t o the shea r deformation o f
rubber bearings , therefor e onl y rubbe r bearings defor m laterall y with piers
at the stabl e position . Th e pea k frequenc y o f th e pie r top i s 2.59H z whic h
is th e eige n value o f th e pier , littl e effec t o f th e superstructur e was
observed, an d only th e pier s deform . I n consequence th e superstructur e and
substructures behav e rathe r independently , an d the bendin g momen t a t the
pier botto m i s les s tha n static bending moment .
A coupl e o f experiment s wer e carrie d out t o investigat e th e basi c
behaviors an d strength o f rubbe r bearings du e t o larg e lateral forces. Fig . 13
shows th e horizonta l forc e an d displacement relatio n o f th e high dampin g
rubber bearin g (HDRB) . Histerisi s wa s observe d durin g cycli c motions .
Picture 1 shows th e deformatio n o f th e bearin g with the lateral/vertical
displacement rati o r of 220% . N o damag e i s occurre d during thes e tests .
Fig.8 P4 -P12 Piers (unit:mm)
Design of the Trans-Tokyo Bay Bridge
(a) Lead Rubber Bearing (b) High Damping Rubber Bearing
(c) Ring Rubber Bearing
F i g . 9 Rubber Bearings
<8>
/777 w 7
S7 -7 /7T
F i g . 1 0 Dynamic Model
(Argirder, Brpier top, Crpier bottom)
400 r
200
<0
-200
Time(sec)
-400
JI I II I L
10 1 21 41 6
02 4 6
F i g . 1 1 Design Seismic wave

Shioi and Nakamura
Picture 1 Lateral Deformation of HDRB

Design of the Trans-Tokyo Bay Bridge 8 3
On one o f th e specime n surfac e a 10c m cut was mad e purposely , an d the

specimen wa s loade d verticall y an d laterally wit h r of 220% , bu t th e cu t di d
not develop .
5. Expansio n joint s
Table 1 shows th e longitudina l movement s o f girde r at P3 and P13 due to

earthquake an d temperature change. Th e desig n movement whic h
expansion join t ca n accommodate i s decide d t o absor b th e 110 % o f
calculated displacement s plu s 30m m t o includ e additiona l movement s whic h
may resul t fro m constructiona l errors , girde r rotation an d minimum gap , and
reaches mor e tha n lm .
Table 1 Longitudinal movement of expansion joints
P3 P12
(P2) (P3) (Pll) (P13)
EARTHQUAKE 240 139 158 272
T E M P E R A T U R E (35deq) 82 377 312 197
EXPANSION J O I N T S 1030 mm 1155 mm
The cantileve r com b joint show n in fig . 14 i s favore d t o accommodate

such a long longitudina l displacements . Th e joint i s mad e o f cas t iron,
which has a good anti-corrosio n property , an d is bolte d t o the uppe r flang e
of th e girder . A s th e structura l form o f thi s join t i s simple , th e replacemen t
is relativel y eas y an d noise complain t du e t o vehicles ma y no t exist .
84 Shi oi an d Nakam ura
The devic e t o preven t fallin g th e girde r fro m pier s i n cas e o f earthquak e

is attache d a t the girde r ends to connec t th e bot h girder s a s shown in fig . 15.
This devic e i s designe d to wor k afte r th e collaps e o f expansio n joint s an d
bearing stoppers .
Fig.15 B r i d g e Falling Prevention Device
6. Conclusio n
The structura l form s o f th e ver y lon g continuou s girde r o f th e TTB

Bridge i s introduce d i n thi s paper . A t presen t 6 piers hav e bee n already
completed an d th e res t pier s ar e now unde r construction . Th e
superstructures ar e now bein g fabricated a t factories. Th e whole highway
will b e finishe d an d open by 1996 .
The author s greatl y appreciat e Prof . M . It o o f Saitam a Universit y an d Mr.

H. Iked a o f Japa n Highwa y Corporatio n fo r givin g them th e opportunit y t o
participate i n thi s colloquium.
9 C O N T I N U O U S S T E E L AN D
COMPOSITE BRIDGE S
J.D. PARSON S
Cass Hayward & Partners, Chepstow, UK
Abstract
This paper describes how the problems associated with expansion joints
and bridge bearings in simply supported spans of multi-span steel
structures can be overcome by the preferred and significantly more
durable continuous steel and composite bridge construction. Totall y
jointless steel and composite bridges are discussed and integral abut-
ments illustrated. Achieveabl e continuty, economy and aesthetics are
also considered. Retrofi t of existing bridge structures is mentioned.
Keywords: Continuou s Steel and Composite Bridges, Durable, Economic,
Aesthetic.
1 Introduction
Designers of steel and composite bridges have increasingly been moving

towards joint-free continuous bridges for many years. I n adddition to
the reduction in maintenance liability associated with moving parts,
i.e. expansion joints and bearings, continuous steel and composite
bridges have other advantages such as economy both for the super and
substructure and improved aesthetics.
Having recognised the problems associated with joints and bearings
designers are giving attention to the ends of bridges, both single and
multi-span, where inspection and maintenance galleries are provided
often complicating construction of the abutments. Th e way forward
appears to include the consideration of integral abutments connected
to the deck which deletes joints and bearings altogether, removes the
need for galleries and simplifies once again construction of the
abutments.
Rehabilitation of existing steel multi-span structures is
undertaken in other countries where a no-joint or minimal-joint policy
already exists. T o date there is little experience of such retrofit
in U.K. but as strengthening of the bridge stock continues this may be
the way forward for some bridges.
2 Continuity in the superstructure
2.1 The problem

The problems associated with expansion joints and bearings on simple
Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published in 1 9 9 4

86 Parsons
bridges, including those of steel or composite construction, either

single span or multi-span are well known. Maintenanc e and repair
expenses have risen sharply in recent years and not insignificant sums
have been spent on these areas.
Damage can be partially attributed to the increase in traffic
volume, especially of heavy vehicles. Impac t forces caused by
vehicles passing over the expansion joints, is alone often sufficient
to lead to problems. Unfortunatel y impact is often combined with poor
detailing. Eventuall y water and de-icing salts penetrate through
joints and leaks down onto the superstructure and substructure below.
Damage occurs to the expansion joint and to the reinforced concrete
slab ends. End s of girders and steel bearings corrode and
deterioration leading to unsightly paintwork and staining of the
substructure. Reinforce d concrete crossheads and piers also corrode
following attack by chloride penetration.
Access for inspection to these vulnerable parts is usually
difficult and repairs, which nearly always have to be performed while
the bridge is in use, and maintenance, already restricted in many
ways, tend to be further complicated.
A feature of simple multi-span bridges is the poor ride quality of
the numerous and other frequent expansion joints (a notable example is
the Midland Links).
In BS.5400:Part 1 0 Code of Practice for fatigue, allowance for
impact is to be made. Wher e a discontinuity occurs in the road
surface, e.g. at an expansion joint, the static stress at every point
affected by a wheel, at or within 5 m of the discontinuity, should be
increased by magnifying the relevant influence line by 25% at the
discontinuity reducing to zero allowance at 5m .
2.2 The solution

Since the 1970 's most multiple span steel bridges have been made
continuous both for dead loads and live loads over intermediate
supports. Thi s reversed the practice up to 197 0 reported by Holland;
of 2 8 viaducts put out to contract between 196 7 and 196 9 only 14 1
(15»1%) o f 93 3 spans were continuous form fo r multi-span structures.
Experience is that continuous bridges decks are significantly more
durable then those consisting of simply supported spans. Th e
Department of Transport strongly recommends the use of continuous
decks wherever possible.
2.3 The practice

Spans are usually fixed by site restrictions or obstacles and required
clearances. Wher e freedom exists an economical span in the range 25 m
to 45 m is likely. However , where deep piled or other expensive
foundations are needed, budget costing will encourage the use of
longer spans.
Multiple spans of approximately 26 m suit universal beams. Th e
optimum for using plate girders for continuous bridges is about 45m ,
because 27 m long "span girders" which should be a lighter section, can
be spliced with "pier girders" of a single plate l8 m long. Fo r longer
spans, more shop or site splices are needed. Componen t lengths for
shop fabrication should be the maximum possible consistent with
delivery and site restrictions to reduce the amount of on-site
Continuous steel and composite bridges 8 7
assembly. Th e maximum length for road delivery without restrictions

in the U.K., is normally 27.4 m although longer lengths can be
transported by arrangement.
Generally the girders run over the piers and are site spliced
within the span near the points of contraflexure. Sit e splices may be
formed with high strength friction grip bolted joints. Sit e welded
joints are sometimes used by erection is less rapid.
Ideally the length of end spans should be about 0. 8 of the
penultimate span. However , because of site conditions continuous
bridges are often curved in plan, have irregular span, or skews. Wit h
these structures there can be a tendency for negative reactions which
can further exacerbate the damage at bearings or supports due to
impact, vibrations or reversal of stresses.
In continuous bridges with integral steel crossheads the concrete
deck slab at intermediate supports is likely to be more heavily
reinforced due to the combination of longitudinal global and local
bending. Th e control of crack widths here, important for long term
durability, may govern the reinforcement design.
3 Continuity with the substructure
3.1 Integral steel crossheads

Earlier composite steel bridges tended to use closely spaced multiple
girders (1.5 m to 3.5 m spacing) with leaf piers supporting every girder
on many bearings. Fo r wide decks such piers are wasteful and
aesthetically undesirable so that from the 1970 's column type piers
Fig. 1a Multiple U B deck on leaf pier
Fig. 1b Multiple U B deck with integral steel crosshead

88 Parsons
surmounted by steel crossheads integral with the deck have often been
used. Fewe r bearings are demanded which with a crosshead shallower
then the main girder can be raised into a more sheltered environment
(see fig.1).
Examples are the M27 River Hamble Bridge, M25 Mar Dyke Viaducts,
Skipton Bypass, M25 Brook Street Viaduct and M25 /A1(M) South Mimms
Interchange Bridges. A recent example is the Blaydon Haughs Viaduct
(Al Newcastle Western Bypass) which won a Steel Award in 1991.
3*2 Flexible piers

When intermediate supports are tall or can be slender then moving
bearings can be avoided altogether. Th e movements caused by thermal
expansion and contraction are accommodated by bending of the piers.
Other design load effects such as longitudinal braking or traction can
be shared between the piers in proportion to the pier stiffness.
Examples are the M 5 Avonmouth Bridge where flexible piers 1.2 m x
4.27m range in height from 13 m to 33 m and bearings are the fabricated
knuckle type, and the recently constructed pair of 3 span River Avon
Bridges (Bristol Spine Road) where fixed bearings are employed at each
of the intermediate piers (1.7 m x 1.0 m and 7 m high).
3.3 Integral abutments

Completely jointless steel bridges, i.e. continuous span bridges
without expansion joints at either end due to the connection of the
superstructure to the abutment, are rare in the U.K. but have been
adopted for a number of bridges in the U.S.A. Wasserma n (1986 )
reported that jointless bridges had been constructed for the past 2 0
years and that 80% of bridges in Tennessee were being constructed
without expansion joints. Thi s included composite steel bridges of up
to 127 m length. I t was admitted that the practice seemed to defy the
calculations which showed that local damage or cracking should have
occurred at abutments.
In attempting to justify the use of jointless bridges it should not
be forgotten that many older U.K. structures, especially railway
bridges, had no expansion facilities at all, although if multiple
span, they were usually simply supported so that some flexibility was
achieved between one span and another, even if unintentionally.
Girder ends were usually built in at abutments without bearings or
joints. Suc h structures have often experienced no problems and appear
to defy calculation, although most are of short length compared with
modern viaducts.
From the U.S. experience it would appear to be feasible to use
jointless steel deck in the U.K., where the temperature range is
generally less severe. However , conditions may not always be
suitable, as when abutments are comparatively rigid. Th e concept
envisaged by Hambly and Nicholson (1990 ) whereby the abutment slides
on supporting granular fill implies on supporting granular fill
implies movement such that a roadway expansion joint may still be
desirable or regular repair of the surfacing locally will be
necessary. Figur e 2 shows possible abutment details for jointless
steel bridges. Figur e 2 (a) requires the abutment to slide on granular
fill, Figure 2 (b) involves flexure of piles and Figure 2 (c)
incorporates a joint, but which is buried at deck soffit level. Th e
Continuous steel and composite bridges
RUN-ON SLAB MOVEMENT
MOVEMENT BY
SLIDING O N
Fig.2a G R A N U L A R FILL
MOVEMENT
MOVEMENT BY
PILE FLEXURE
Fig.2b
MOVEMENT
Fig.2c RIGID A B U T M E N T
Fig.2. Abutment details for jointless steel bridges

90 Parsons
arrangement of 2 (c) has been used on railway bridges with no apparent

difficulty. I t implies that the fill behind the deck must compress to
take up the movement. I n all cases, the steel girder ends are cast
into the abutment or curtain wall which is easily achieveable, but it
is desirable to incorporate a weather flat at the steel/concrete
interface to prevent undetected corrosion of the steel web.
An additional advantage of integral abutments is that the
inspection gallery required for inspection maintenance of expansion
joints and bearings can be removed simplifying once again the
construction of the abutment. However , reinforced concrete run-on
slabs are required to prevent excessive compaction of the fill behind
the abutment.
4 Poor ground and continuity
Traditionally for mining areas or whenever the interpretive soils

report warned of a risk of relative settlements, simple spans were
used. I n the computer we how have the tool to make designing
continuous bridges in such circumstances manageable and the designer
can make due allowance in the design for possible displacements in the
dead load moments and shear effects due to settlements. Th e use of
slender decks reduces the effects of settlements.
5 Aesthetics
A continuous steel or composite bridge gives the designer scope for

producing an elegant structure of clean appearance with pleasing
lines. Welde d site splices can improve appearance for bridges in
sensitive areas. Pair s of stiffeners at supports can be 'boxed ' in.
Integral steel crossheads can be used to reduce the number of piers
and raise the bearing out of sight.
Curved girder soffits also greatly improve the line and interest of
the structure as well as making economic use of often increased
construction depth adjacent to piers. Fig . 3 shows the aesthetic
benefits of a continuous girder bridge compared to a simple girder
structure (as can be seen at the Midland Links) as well as reducing
the number of expansion joints and bearings and reducing size of sub-
structure .
Curved bridges are dealt with by making the girders from straight
chords between site splices, but an increasing U.K. trend for sharper
radii is to use curved girders taking advantage of modern fabrication
methods.
Examples are the Simon de Montfort Bridge, Evesham Bypass (Award
1988) an d Festival Park Flyover, Stoke on Trent (Award 1990) .
Incidentally, both examples are of twin plate girder and cross girder
type form with only two bearings with supporting columns at
intermediate piers.
Continuous steel and composite bridges 9 1
EXPANSION JOINT S
/
Fig.3.a. Simple girder structure

BEARINGS
1/ N
TT n
Fig.3.b. Reduction of number of bearings and expansion joints by
use of aesthetically pleasing continuous curved soffit girders
Fig.3. Continuous girder structure example

6 Erection
In addition to crane erection of full span lengths continuous bridges

can be successfully erected by the launch method. Althoug h this
method requires the girders to have a level soffit site assembly of
the individual girder elements are at ground level and craneage
requirements are significantly reduced.
Examples are Loughor bridge, Llanelli and Chepstow Viaduct (1987 )
with a main span of 93 m and overall length 302.5m .
7 Retrofit continuity
The current programme of bridge strengthening and repairs gives scope

for reducing or improving future maintenance on existing structures.
A suitable rehabilitation method may be based on one used in Australi
for new designs, where continuity for live loads only is achieved for
simply supported steel girders which are concreted in over the piers
with rebars in th e deck slab forming tensile continuity. Thi s could
work well where deteriorated deck slabs necessitate concrete removal.
Experience in Tennessee, U.S.A. is that for short span steel
bridges no girder end modification is required. Fo r longer steel par
compression flanges are jointed. I n some cases, bearings must be
modified. Wher e the foregoing modifications are employed, if
structure widening is to be added, the new portion will be of fully
continuous composite construction.
8 Conclusion
Multi-span steel bridge should always be made continuous because this

92 Parsons
is easier to achieve and is economic. Th e use of integral abutments

to achieve a completely jointless deck for both single and multi-span
bridges should be considered further taking into account the success
of the practice in other countries. Th e economy of repairing the road
surfacing compared with provision of joints requires consideration.
The amount of theoretical movement which can be tolerated without;
joints is tentatively suggested at ±25mm for surfaced carriageways and
±40mm for ballasted rail tracks. Fo r structures fixed at one end this
implies a maximum bridge length of about 70 m for highway bridges and
120m for rail bridges. Thes e lengths can be doubled if the structure
is fixed at mid length.
Continuous steel and composite bridges can be both aesthetically
pleasing and durable for the future. Aestheti s is important, but
maintenance is of greater concern.
9 References
Hambly, E.C. and Nicholson, B.A. (1990) Prestressed beam integral

bridges. Th e Structural Engineer, December
Holland, A.D. and Deuce T.L.G. (1970 ) A Review of Small Span Highway
Bridge Design and Standardisation. Journa l of the Institution of
Highway Engineers, August
Wasserman E.P. (1986 ) Jointless Steel Bridges. Nationa l Engineering
Conference, AISC, Nashville
REPORT O N SESSIO N THRE E
G.P. T I L LY
Dr Nakamura explained tha t the trans Tokyo Bay bridge was designed i n continuous
sections, eac h composed o f 9 to 1 1 spans. Th e crossing includes a suspension bridg e
having a spa n o f 199 0 m.
In respons e t o a questio n abou t ho w th e rubbe r bearings describe d i n th e
paper hav e bee n designe d t o hav e highe r values o f damping , th e autho r explained
that constituent s ha d bee n adde d t o th e rubbe r when i t was manufactured , so that
the resultin g materia l of th e typ e (b ) bearing show n i n Figur e 9 ha d increased the
hysteresis. Th e type (c ) bearing had a steel rin g which strengthened the connectio n
between th e rubbe r and stee l plate . Thi s bearing di d no t hav e hig h damping .
It was commente d tha t experience o f th e tune d dynami c absorber s that are
sometimes use d t o dampe n lively bridges was that they are sensitive t o temperatur e
and there can be problems with durability. I n general, design lives of bridges can be
as hig h a s 10 0 years but i t shoul d b e recognise d tha t joints an d bearings canno t be
expected t o perfor m satisfactoril y for a s lon g a s thi s an d design s shoul d tak e int o
account th e nee d t o fi t replacement s at suitable intervals.
In response t o a questio n abou t composit e construction , Dr Nakamura said
that i n Japanes e bridg e design , composit e stee l girder s ar e no t use d becaus e
designers ar e no t full y satisfie d tha t the y hav e enough informatio n t o cop e wit h
hogging moments . A speake r commente d tha t i n th e Unite d State s thi s i s no t
regarded as a problem. Design s of composite an d non-composite bridges are similar
and no accoun t is take n o f cree p and shrinkage . I n the pas t it ha d been th e usua l
practice t o cas t th e i n sit u concret e in a correc t sequenc e i n orde r t o balanc e th e
loads bu t th e concret e is no w machin e laid an d is place d in a continuous run , from
one en d o f th e bridg e t o th e other , withou t stopping . Transvers e crack s ar e
sometimes develope d i n th e concret e but thi s i s no t regarde d as a problem.
Dr Parson s presented hi s pape r on continuou s stee l an d composite bridges .
He discusse d th e movement o f integral abutments and suggested tha t it occurs by a
sliding mechanism. Durin g th e discussion a speaker questioned whethe r rotation is
more likely to occur . Th e autho r felt tha t the movement wa s relate d to stiffnes s o f
the abutmen t and there could be cree p of the soil . Thi s type o f desig n i s known t o
perform satisfactorily in practice but he was able to provide an adequate explanation
of th e mechanics.
A speake r commente d tha t in Franc e ban k seat s ar e used wit h a singl e lin e
of pile s t o reduc e movemen t capacit y an d permi t rotation . Ther e ar e 15 0 suc h
bridges on the Alpine road s and they have now been performin g satisfactorily for 12
94 Tilly
years. I n Franc e run-o n slab s ar e commonl y use d bu t ar e designe d wit h a join t

between th e bridge deck and the slab. Th e running surface is laid continuously ove r
the joint t o provid e a smoot h ride .
In the UK there were problems with run-on slabs breaking due to settlemen t
and thei r us e wa s abandone d abou t 1 0 years ago. I t is no w require d that they b e
designed t o b e cantilevere d o r supporte d o n piles . I n th e Unite d State s 7 t o 9 m
long run-off slabs ar e commonly used. Th e soi l beneath is compacte d according to
normal requirements and it is not considered necessary to specify th e use of granular
fill. Th e differen t State s follow thei r own practice .
PART FOU R
INTEGRAL
BRIDGES
10 D A N I S H I N T E G R A L
BRIDGES
E.M. VEJ E
Cowiconsult, Consulting Engineers an d
Planners AS , Lyngby, Denmar k
Abstract
This p a p e r g i v e s a d e s c r i p t i o n o f th e r e c e n t majo r Danis h b r i d g e p r o -
j e c t s , e s p e c i a l l y th e Grea t B e l t Lin k P r o j e c t , focusin g o n desig n
d e v e l o p m e n t s fro m t h e L i t t l e B e l t Suspensio n B r i d g e , opene d f o r t r a f f i c
in 1 9 7 0 , t o Grea t Bel t Eas t Bridge , ' s e t fo r construction * completio n
1 9 9 7 / 9 8 w i t h s p e c i a l a t t e n t i o n t o t h e e f f o r t s mad e t o m a x i m i z e t h e
length o f th e continuou s superstructur e sections .
Keywords: F a r 0 B r i d g e s , G r e a t B e l t Link , Expansio n J o i n t s ,

D e h u m i d i f i c a t i o n s y s t e m , C l o s e d Bo x G i r d e r s
1 Introductio n
I n r e c e n t y e a r s d e s i g n s o f m u l t i - s p a n b r i d g e s u p e r s t r u c t u r e s sho w a
c l e a r t r e n d toward s long , continuou s s e c t i o n s . Thi s i s i n acknowledge -
m e n t o f t h e b e n e f i t s o f h i g h d r i v i n g c o m f o r t an d r e d u c e d m a i n t e n a n c e
e f f o r t whe n t h e numbe r o f e x p a n s i o n j o i n t s i n t h e s t r u c t u r e an d t h e
roadway pavemen t i s l i m i t e d .
This d e s i g n t r e n d i s d r i v i n g th e developmen t o f expansio n j o i n t
s t r u c t u r e s a n d b r i d g e b e a r i n g s t o w a r d s i n c r e a s e d movemen t c a p a c i t y .
V i c e v e r s a , t h e c a p a c i t y o f t h e s e s p e c i a l component s s e t s l i m i t s t o th e
possible continuit y lengt h o f th e s t r u c t u r e s .
2 Th e Far0 B r i d g e s
The F a r 0 B r i d g e s c o m p r i s e tw o b r i d g e s , a n o r t h e r n b r i d g e f r o m Z e a l a n d
t o F a r 0 a n d a s o u t h e r n b r i d g e f r o m F a r 0 t o F a l s t e r . Th e n o r t h e r n b r i d g e
c o m p r i s e s 1 8 s p a n s o f 8 0 m l e n g t h , p l u s tw o en d s p a n s o f 7 8 m . Th e
southern bridg e c r o s s e s th e navigatio n channe l b y a cabl e staye d bridg e
s i t u a t e d a p p r o x i m a t e l y midway . I t h a s a c e n t r a l spa n o f 2 9 0 m f o r s h i p
p a s s a g e an d s i d e span s o f 1 2 0 m e a c h . A t t h e end s o f t h e c a b l e s t a y e d
b r i d g e 1 5 approac h span s w i t h th e s e l e c t e d s t a n d a r d l e n g t h o f 8 0 m
( e x c e p t f o r t h e tw o en d s p a n s ) c o m p l e t e t h e b r i d g e ( 7 an d 8 s p a n s
r e s p e c t i v e l y ) . Th e b r i d g e wa s o p e n e d f o r t r a f f i c i n 1 9 8 3 .
The s u p e r s t r u c t u r e i s o u t l i n e d a s a c l o s e d s t e e l b o x g i r d e r , w i t h
a l m o s t t h e sam e o u t e r s h a p e o n t h e e n t i r e b r i d g e l e n g t h . Th e c l o s e d
s t e e l b o x g i r d e r c o n c e p t , w h i c h a l s o wa s a p p l i e d o n t h e L i t t l e B e l t

98 Veje
Zealand Faro
II | i ,H l
~n\ 7i \ 71 \ /f t
Falster
Faro
78m H-120m-H -1-120™-+-
Figure 1 : Th e Z e a l a n d - F a r 0 a n d F a r 0 - F a l s t e r B r i d g e s
Suspension Bridge , b u i l t 1 5 year s e a r l i e r , contain s severa l feature s

fundamental f o r c o s t e f f i c i e n c y :
g r e a t u n i f o r m i t y o f t h e i n d i v i d u a l element s o f th e bo x g i r d e r ,
thereby f a c i l i t a t i n g industrialize d fabricatio n
s i n g l e c e l l bo x g i r d e r c r o s s s e c t i o n w i t h o u t t r a d i t i o n a l t r a n s -
v e r s e an d l o n g i t u d i n a l g i r d e r s
simple c o n n e c t i o n d e t a i l s betwee n t r a n s v e r s e bulkhead s (pe r 4 m )
and t h e e x t e r i o r g i r d e r p a n e l s , t h e r e b y r e d u c i n g l a b o u r consumin g
assembly t o l e r a n c e problem s
c o r r o s i o n p r o t e c t i o n b y mean s o f d e h u m i d i f i c a t i o n o f t h e i n t e r i o r
and a smooth , m i n i m i z e d o u t e r s u r f a c e a r e a t o b e p a i n t e d ; a c o n -
cept s u c c e s s f u l l y introduce d o n th e L i t t l e Bel t Suspensio n Bridg e
Another c h a r a c t e r i s t i c fo r th e Far 0 bridge s i s larg e continuou s j o i n t -

l e s s l e n g t h o f t h e b r i d g e s - p r o b a b l y l a r g e r t h a n f o r an y o t h e r b r i d g e
a t t h e t i m e o f c o n s t r u c t i o n . Th e tw o b r i d g e s a r e c o n t i n u o u s f r o m a b u t
18600
6 0 0 , 100 Q 2x3500 100Q1300,
r 2 Lane s 11 1
| \J W \J \J u w
r\ r\ r\ r\ <r\ r\ r\ r\ / A r\ I
Figure 2 : Close d s t e e l bo x g i r d e r
Danish integral bridges 9 9
ment t o a b u t m e n t , c o r r e s p o n d i n g t o e x p a n s i o n s e c t i o n l e n g t h s o f 1 5 9 6 m
a n d 1 7 2 6 m r e s p e c t i v e l y , an d h a v e o n l y e x p a n s i o n j o i n t s a t t h e a b u t -
ments .
A continuou s s t r u c t u r e i s favourabl e i n severa l r e s p e c t s :
many s e r v i c e an d m a i n t e n a n c e p r o b l e m s a r e e l i m i n a t e d ,
t r a f f i c c o m f o r t i s improve d remarkably ,
c o n s t r u c t i o n cost s ar e reduced .
The l a r g e e x p a n s i o n s e c t i o n s i m p l y l a r g e m o v e m e n t s , w h i c h a l s o
i n c r e a s e th e wea r o f th e b e a r i n g s . Fo r t h a t reaso n th e accumulate d
t r a v e l a t e a c h a b u t m e n t i s m e a s u r e d b y mean s o f d i s p l a c e m e n t t r a n s -
d u c e r s . Thi s g i v e s u s e f u l informatio n abou t th e wea r o f th e b e a r i n g s
and i s o f i m p o r t a n c e f o r t h e d e t e r m i n a t i o n o f t h e i n s p e c t i o n i n t e r v a l s .
The e x t r e m e m o v e m e n t s a t t h e b r i d g e e n d s amoun t t o u p t o 1 0 5 0 m m p e r
j o i n t , primaril y cause d b y temperatur e v a r i a t i o n s .
With r e f e r e n c e t o th e p r o p e r functionin g o f th e bridg e th e c h o i c e o f
a p p r o p r i a t e b e a r i n g t y p e s become s a m a t t e r o f g r e a t i m p o r t a n c e unde r
s u c h e x t r e m e c o n d i t i o n s . Th e b e a r i n g s m u s t b e r e l i a b l e , s o l i d an d
d u r a b l e an d t h e i r r e s i s t a n c e t o movement s mus t b e s m a l l , i f a p r o b l e m -
f r e e f u n c t i o n i n g i s t o b e a t t a i n e d t h r o u g h o u t a lon g spa n o f t i m e .
The i n s t a l l a t i o n o f t h e b e a r i n g s s h a l l b e p l a c e d v e r y a c c u r a t e ,
h e r e u n d e r t h e a l i g n m e n t o f t h e b e a r i n g s an d s i d e g u i d e s .
The f i x e d b e a r i n g s a b s o r b t h e l o n g i t u d i n a l f o r c e s du e t o b r a k i n g an d
wind l o a d s . Mos t i m p o r t a n t f o r t h e g i r d e r a r e t h e l o n g i t u d i n a l f o r c e s
t h a t m i g h t r e s u l t f r o m u n e q u a l f r i c t i o n i n t h e b e a r i n g s o f t h e tw o
bridge h a l v e s . Experience s gaine d s o f a r regardin g th e accumulatio n o f
f r i c t i o n a l f o r c e s o v e r a g r e a t numbe r o f b e a r i n g s ma y b e u s e f u l f o r t h e
p o s s i b i l i t y o f allowin g longe r bridg e section s t o b e buil t withou t
j o i n t s . Th e o b s e r v a t i o n s a r e e n v i s a g e d t o b e c o n t i n u e d o v e r a p e r i o d o f
2 0 - 2 5 y e a r s , wit h a s t a t u s ever y 5t h y e a r , s o t h a t a n adequat e basi s
f o r e v a l u a t i o n ma y b e e s t a b l i s h e d b y e x p e r i e n c e .
S p e c i a l s p r i n g s hav e bee n a r r a n g e d a t th e f i x e d b e a r i n g a l l o w i n g f o r
m e a s u r i n g t h e h o r i z o n t a l f o r c e a c t i n g , p r i m a r i l y o r i g i n a t e d fro m a c t i v e
f r i c t i o n an d f r i c t i o n d i f f e r e n c e s . M e a s u r e m e n t s s o f a r p e r f o r m e d h a v e
r e v e a l e d l o n g i t u d i n a l f o r c e s i n th e orde r o f 30 2 o f th e expecte d f o r c e s
a n d a m o u n t t o l e s s t h a n 10% o f t h e maximu m f o r c e t h a t c a n b e
r e s t r a i n e d . I n t h e d e s i g n i t h a s bee n assume d t h a t t h e d i f f e r e n c e i n
r e s i s t a n c e b e t w e e n t h e b e a r i n g s ma y b e 2% o f t h e v e r t i c a l l o a d .
3 Th e Grea t B e l t Lin k
The 1 8 k m w i d e G r e a t B e l t i s d i v i d e d i n t o tw o m a i n c h a n n e l s , e a s t an d
w e s t , b y t h e t i n y i s l a n d o f S p r o g 0 , T h e r e f o r e , i t wa s o b v i o u s t o t a k e
advantage o f thi s islan d a s a n integrate d par t o f a fixe d lin k p r o j e c t ,
a ' s t e p p i n g s t o n e ' b e t w e e n e a s t a n d w e s t a c r o s s t h e w a t e r . Th e l i n k
c o n s i s t o f t h r e e m a j o r s t r u c t u r e s : A b o r e d r a i l w a y t u n n e l an d a h i g h
l e v e l m o t o r w a y b r i d g e a c r o s s t h e e a s t e r n c h a n n e l , an d a lo w l e v e l d u a l
mode b r i d g e f o r r a i l w a y an d m o t o r w a y a c r o s s t h e w e s t e r n c h a n n e l .
100 Veje
3 . 1 Th e Wes t B r i d g e
The 6 . 6 k m b r i d g e c o n s i s t s o f tw o h a u n c h e d c o n c r e t e b o x g i r d e r s w i t h
v a r i a b l e d e p t h e a c h s u p p o r t e d o n s e p a r a t e p i e r s h a f t s s h a r i n g a commo n
s u b s t r u c t u r e d e s i g n e d a s a g r a v i t y f o u n d e d c a i s s o n . Th e n o r t h e r n g i r d e r
c a r r i e s t h e r a i l w a y t r a c k , an d
the souther n th e roa d t r a f f i c .
The r a i l w a y g i r d e r h a s a n o v e r a l l
d e p t h o f 8 . 7 0 m , an d 5 . 1 3 m a t
mid s p a n . Th e r o a d w a y g i r d e r
d e p t h s a r e 7 . 3 4 m an d 3 . 7 8 m r e -
spectively.
The s u p e r s t r u c t u r e i s d i v i d e d
i n t o 5 1 mai n span s o f 1 1 0 . 4 0 m
and 1 2 e x p a n s i o n j o i n t span s o f
8 1 . 7 5 m . Expansio n j o i n t s ar e
p r o v i d e d a t t h e a b u t m e n t s an d a t
4 intermediat e p i e r s , thu s subdi -
F i g u r e 4 : Th e W e s t B r i d g e viding th e o v e r a l l lengt h int o
s i x continuou s g i r d e r s o f abou t
1100 m .
The r a i l w a y e x p a n s i o n j o i n t i s a n e s s e n t i a l p a r t w i t h r e g a r d t o t h e
s a f e o p e r a t i o n o f t h e r a i l w a y . Th e j o i n t u s e d o n t h e W e s t B r i d g e i s a
new d e v e l o p m e n t a n d i t h a s b e e n d e c i d e d t o i n s t a l l an d t e s t a p r o t o t y p e
o f t h e r a i l e x p a n s i o n j o i n t i n t h e Danis h S t a t e R a i l w a y s ' mai n l i n e .
The e x p a n s i o n j o i n t s c o n n e c t i n g t h e c o n t i n u o u s s u p e r s t r u c t u r e o f
a b o u t 1 1 0 0 m l e n g t h w i l l mov e b e t w e e n ± 3 5 0 m m an d ± 6 0 0 mm . A t o t a l
c a p a c i t y o f u p t o 1 2 0 0m m h a s s o f a r onl y bee n see n i n J a p a n f o r t h e
Honshu-Shikoku B r i d g e p r o j e c t .
The r e q u i r e m e n t s f o r a n g u l a r r o t a t i o n , v e r t i c a l r a d i u s o f t h e r a i l
and c o m f o r t l e v e l a r e p a r a l l e l r e q u i r e m e n t s f o c u s i n g o n a h i g h c o m f o r t
l e v e l a n d h i g h s a f e t y o f o p e r a t i o n . Th e r a i l w a y e x p a n s i o n j o i n t a r r a n -
g e m e n t s i s c a p a b l e o f m e e t i n g t h e s e r e q u i r e m e n t s , an d b a s e d o n d e v e l o p -
m e n t s o f e x p e r i e n c e g a i n e d i n J a p a n an d G e r m a n y .
3.2 Eas t B r i d g e , Approac h Bridg e
The E a s t B r i d g e i s d e s i g n e d w i t h e x t r e m e l y l a r g e j o i n t l e s s s e c t i o n s ,
p r o b a b l y t h e l o n g e s t e v e r p e r f o r m e d i n b r i d g e d e s i g n . Th e e n t i r e 6 . 8 k m
b r i d g e i s o n l y s u b d i v i d e d i n t o 3 e x p a n s i o n s e c t i o n s o f 2 . 7 km , 2 . 5 k m
a n d 1 . 5 km .
Figure5 : Th e E a s t B r i d g e E l e v a t i o n
The tw o a p p r o a c h b r i d g e s a r e 1 5 3 8 m an d 2 5 3 0 m l o n g r e s p e c t i v e l y ,
w i t h i n t e r m e d i a t e s p a n l e n g t h s o f 1 9 3 m an d e n d s p a n s o f 1 4 0 m an d 6 2
m. Th e a p p r o a c h b r i d g e g i r d e r i s b a s e d o n c l o s e d b o x s e c t i o n s w i t h a
c e n t r a l l o n g i t u d i n a l b u l k h e a d w h i c h m i n i m i z e t h e s h e a r l a g e f f e c t an d
e n s u r e s s h e a r c a p a c i t y i n c a s e o f s h i p c o l l i s i o n a g a i n s t on e o f t h e
i n c l i n e d s i d e s . Th e i n t e r i o r o f t h e g i r d e r s ( b o t h a p p r o a c h an d s u s p e n -
s i o n b r i d g e ) i s u n p a i n t e d an d p r o t e c t e d b y d e h u m i d i f i c a t i o n a s f o r t h e
F a r 0 b r i d g e s . Abou t 8 0 % o f t h e t o t a l s t e e l s u r f a c e i s thu s u n p a i n t e d
w i t h s u b s t a n t i a l c o s t a n d t i m e s a v i n g s a s a r e s u l t . Th e d e p t h o f t h e
girder i s 6. 7 m .
E a c h a p p r o a c h b r i d g e i s c o n t i n u o u s fro m t h e abutmen t t o t h e a n c h o r
b l o c k o f t h e s u s p e n s i o n b r i d g e . Th e c o r r e s p o n d i n g ma x t o t a l movemen t o f
t h e e x p a n s i o n j o i n t s i s f r o m 7 9 0 m m u p t o 1 5 3 0 mm . W a t e r t i g h t m o d u l a r
e x p a n s i o n j o i n t s o f t h e s o - c a l l e d Swive l J o i n t typ e hav e bee n s e l e c t e d
as a r e s u l t o f i n t e n s i v e t e s t s t o achiev e th e require d 10 0 y e a r f a t i g u e
l i f e . T o suppor t th e r e l a t i v e l y f l e x i b l e s t e e l s u p e r s t r u c t u r e o n th e
p i e r s , s p h e r i c a l b e a r i n g s h a v e b e e n c h o s e n . Th e a d v a n t a g e o f t h e s e
b e a r i n g s i s t h e r e l a t i v e l y l o w r e s t r a i n i n g moment s o r i g i n a t i n g f r o m
r o t a t i o n f r i c t i o n i n th e s p h e r i c a l p a r t s , whic h i s o f importanc e f o r
t h e d e s i g n o f t h e s u p p o r t i n g s t r u c t u r e s i n t h e b r i d g e g i r d e r . Th e
h o r i z o n t a l movement s a r e t r a n s f e r r e d b y s l i d i n g o v e r a s p h e r i c a l s e g -
m e n t . Th e s l i d i n g mediu m i s PFT E ( T e f l o n ) a g a i n s t s t a i n l e s s s t e e l .
102 Veje
3.3 Eas t B r i d g e , Mai n B r i d g e
The n a v i g a t i o n s p a n o f t h e E a s t B r i d g e c r o s s e s t h e i n t e r n a t i o n a l s h i p -
p i n g r o u t e b e t w e e n t h e B a l t i c S e a an d t h e N o r t h S e a . B a s e d o n i n t e n s i v e
s t u d i e s o f c a b l e s t a y e d an d s u s p e n s i o n b r i d g e c o n c e p t s a s w e l l a s
n a v i g a t i o n a l r i s k s a s u s p e n s i o n b r i d g e w i t h a 1 6 2 4 m m a i n s p a n an d tw o
s i d e s p a n s o f e a c h 5 3 5 m wa s d e c i d e d u p o n .
The m a i n s p a n i s o u t -
l i n e d w i t h a c a b l e sa g
corresponding t o 1/ 9 time s
t h e spa n l e n g t h . V e r t i c a l
hangers eac h 2 4 m suppor t
t h e g i r d e r . Th e s u s p e n s i o n
bridge girde r i s outline d
as a c l o s e d bo x g i r d e r a s
for th e approac h bridge ,
w i t h d e p t h o f 4 m an d
w i d t h o f 3 1 m . Th e b r i d g e
is arrange d wit h a n inno -
v a t i v e a r t i c u l a t i o n aroun d
the pylon s allowin g th e
g i r d e r t o b e continuou s
over th e f u l l lengt h o f
2 . 7 k m b e t w e e n t h e tw o
a n c h o r b l o c k s . Th e t r a d i -
t i o n a l expansio n j o i n t s a t F i u
& r e6 1
A r c h i t e c t ' s vie w o f th e Eas t
the pylon s ar e avoided . Bridg e
Compared t o a s y s t e m w i t h
j o i n t s a t t h e p y l o n s , a n a l y s e s h a v e i n d i c a t e d a n a p p r o x i m a t e l y 25%
r e d u c t i o n i n t h e l o n g i t u d i n a l d e f l e c t i o n o f t h e g i r d e r fro m t r a f f i c
load.
I f f r e e m o v e m e n t s w e r e a l l o w e d , t h e e x t r e m e h o r i z o n t a l movemen t f r o m
t h e c h a r a c t e r i s t i c t r a f f i c l o a d a t t h e e x p a n s i o n j o i n t s woul d b e 1 . 8 m .
In o r d e r t o l i m i t l o n g i t u d i n a l movements , h y d r a u l i c b u f f e r s a r e
a r r a n g e d b e t w e e n t h e a n c h o r b l o c k an d t h e g i r d e r . Th e b u f f e r s a l l o w f o r
s l o w h o r i z o n t a l m o v e m e n t s , f o r i n s t a n c e du e t o t e m p e r a t u r e , u p t o ± 1 . 0
m an d a f r e e r o t a t i o n o f t h e g i r d e r , w h i l e f a s t movements , f o r i n s t a n c e
due t o a h e a v y v e h i c l e p a s s i n g t h e b r i d g e , a r e r e s t r a i n e d i n o r d e r t o
i n c r e a s e t h e f a t i g u e l i f e an d t o r e d u c e t h e w e a r o n t h e e x p a n s i o n
j o i n t . Th e r e m a i n i n g p o r t i o n o f t h e t h e o r e t i c a l movemen t w i l l b e b a l -
anced b y r e s t r a i n i n g f o r c e s i n th e g i r d e r .
The a d v a n t a g e s o f t h e c o n t i n u o u s g i r d e r c o n c e p t a r e :
A v e r y simpl e arrangemen t a t th e pylons .

Reduced i n s t a l l a t i o n an d m a i n t e n a n c e c o s t s f o r t h e e x p a n s i o n
j o i n t s a t th e ancho r block s
Improved s t i f f n e s s o f th e o v e r a l l suspensio n b r i d g e system .
Improved a e r o d y n a m i c a l s t a b i l i t y .
The c o n t i n u o u s g i r d e r c o n c e p t a l s o p e r m i t s e l i m i n a t i o n o f t h e t r a d i -
t i o n a l c r o s s beam s o n t h e p y l o n s i m m e d i a t e l y b e l o w t h e g i r d e r , an d t h u s
c o n t r i b u t e s t o th e a e s t h e t i c a l appearanc e o f th e bridge .
4 Reference s
H a a s , G . P e t e r s e n , A , O s t e n f e l d , K . H . : Th e F a r 0 B r i d g e s , Ne w
D e v e l o p m e n t s i n D e s i g n an d C o n s t r u c t i o n , I n t e r n a t i o n a l Symposiu m o n
S t r a i t C r o s s i n g s , S t a v a n g e r , Norway , O c t o b e r 1 9 8 6 . C O W I - p u b l i c a t i o n
no 5 5 6 .
J a c o b s e n , A . S . P e t e r s e n , A S0rensen , 0 : Desig n o f th e 3 k m Continuou s
S t e e l B r i d g e f o r t h e G r e a t B e l t E a s t B r i d g e , 1 0 t h Annua l I n t e r n a -
t i o n a l B r i d g e C o n f e r e n c e , P i t t s b u r g h , P e n s y l v a n i a , J u n e 1 9 9 3 . COWI -
publication n o 869 .
O s t e n f e l d , K . H . : I m p o r t a n t S t r u c t u r e s an d t h e i r S p e c i f i c P r o b l e m s -
G r e a t B e l t L i n k , Denmark , S e m i n a r o n R e s e n t D e v e l o p m e n t s i n B r i d g e
D e s i g n a n d C o n s t r u c t i o n s , A t h e n s , G r e e c e , Ma y 1 9 9 1 . C O W I - p u b l i c a t i o n
no 7 5 6 .
O s t e n f e l d , K.H . : B r i d g e s o f Toda y an d Tomorrow , 1 0 t h Annua l I n t e r n a -
t i o n a l B r i d g e C o n f e r e n c e P i t t s b u r g h , P e n n s y l v a n i a , J u n e 1 9 9 3 . COWI -
publication n o 868 .
11 P R E C A S T S E G M E N T A L
CONSTRUCTION
J. MULLE R
Jean Muller International, Montigny le Bretonneux,
France
Abstract
This overvie w o f precas t concrete segmental constructio n includes: larg e urban
highway o r mass transi t structures; a composit e (stee l an d concrete ) space trus s
segmentally buil t fo r mediu m spa n viaduct s o r very long cable-staye d spans ; a
special launchin g gantr y application i n long-spa n girde r bridges an d th e
monitoring o f expansio n joint s i n th e firs t attemp t a t intelligent structures ; three
current concret e cable-staye d bridges ; th e ne w bi-staye d concep t allowin g clear
spans o f u p t o 2,40 0 m. Th e followin g paragraph s describe som e salien t feature s
of curren t specific project s a s a n outlin e o f th e method' s stat e o f th e ar t while
projecting it s potentia l development .
Keywords: Precas t segmental construction ; railway viaducts; highwa y bridges ;
composite construction ; long-span girde r bridges; cable-staye d bridges ; spac e
trusses; bi-stayed bridges .
This paper is based on the paper given at the FIP Symposium, Budapest, May 1992
1 Brie f survey o f pas t experienc e
Precast concret e segmental constructio n ha s no w ove r 2 5 years o f experience .

After th e precas t arc h bridges designe d by Eugen e Freyssine t an d built b y
Campenon Bernar d between 194 6 and 1950 , the world' s firs t precas t concrete
bridge wit h match-cast joints, designe d by th e author , was buil t nea r Buffalo, NY ,
in th e US A i n 1952 . Choisy-le-Ro i bridg e nea r Paris over th e Rive r Seine ,
designed an d built i n 1962-6 3 was the n th e firs t applicatio n o f epox y resi n joint s
in match-cas t construction an d represente d th e actua l start o f precas t segmenta l
construction o n a n industria l scale .
Many bridge structure s followed an d were th e occasio n o f successiv e
innovations an d developments . Significan t step s ar e as follows:
1 Match-cas t dry joints - New York , 195 2

2 Epox y match-cast joints - Choisy-le-Roi , 196 2
3 Lon g be d precastin g - Choisy-le-Roi , 196 2

106 Muller
4 Continuou s dec k buil t i n cantileve r - Choisy-le-Roi , 196 2

5 Flexible-walle d pier s - Choisy-le-Roi , 1962 ; Juvisy, 1965
6 I-shape d pier s an d double-walle d pier s - St Andr e d e Cubzac , 1977;
Magnan, 197 3
7 Shor t be d precastin g - Pierre Benite , 1964 ; Paris Ring Roa d Bridges , 196 5
8 Bea m an d winch placin g - Pierre Benite , 196 4
9 Launchin g gantr y - Oleron , 196 4
10 Short-be d precastin g with variable geometr y - Chillon , 196 5
11 Launchin g gantr y with variable geometr y - Chillon , 196 5
12 Progressiv e placin g an d swive l cran e - Rombas, 197 2 Fontenoy, 197 5
13 Multipl e key s i n match-cas t joints - Alpine Motorwa y bridges, 197 1
14 Assembl y o f ful l span s o n falsewor k - Alpine Motorwa y overpasses , 197 1
15 Concret e cable-stayed bridg e with centre suspension - Brotonne, 197 3
16 Precas t segmental concret e urban railroad bridge - Marne-la-Vallee, 197 5
17 Long-spa n segmenta l railroa d bridge - Clichy , 197 6
18 Externa l prestressin g concep t studies , Lon g Ke y Bridge, Florida, USA,
1976 (Fig . 1 )
19 Span-by-spa n constructio n o n assembl y truss ; Transverse prestressed dec k
slab; as-cas t riding surfac e - Long Ke y Bridge, 1976
20 Precas t segmental pier s - Linn Cove , NC , USA , 197 7
2 Recen t development s i n mediu m spa n girde r bridge s
2.1 Sa n Antonio Y Project, TX, USA

Currently, thi s highwa y viaduc t represent s th e singl e largest applicatio n o f precast
segmental constructio n i n a difficul t urba n environment. Th e us e o f a shallow ,
narrow bo x girde r with large overhang s (se e typica l cros s sectio n i n Fig . 2 )
contributes t o th e aesthetic s o f thi s uniqu e project . Externa l prestres s an d span -
by-span constructio n with self-launchin g assembl y trusse s hav e bee n use d t o allo w
maximum constructio n efficienc y an d low cost , Tabl e 1 .
Fig. 1 Lon g Ke y bridge externa l prestress

Precast segmental construction 10 7
13.00
Fig. 2 Sa n Antoni o Y Project , typical cross-sectio n
Table 1 Sa n Antoni o Y Project . Schedul e o f tender s
CONTRACT TENDER DECK ARE A CONTRACT BRIDCE

DATE M2 S F AMOUNT COST
($M) ($/SK)
1A DEC 8 4 34.000 366.000 18 20
2B JUNE 8 5 32.500 350.000 19 25
3A-3B MAY 8 G 52,700 567,000 36 35
IC OCT 8 7 69.700 750,000 43 37
3C-3D AUG 8 9 54.000 581,000 49 47
2C OCT 9 0 40.800 439,000 52 49
TOTAL 283.700 3.053.000 217 37

average
2.2 Monterre y Rail Viaduct , Mexic o

This i s th e singl e largest applicatio n o f precas t segmental constructio n in th e fiel d
of urba n light rai l systems. Th e guidewa y consist s o f a n 18. 5 km long viaduct
with 619 spans o f a n averag e length o f 2 7 m, carryin g twin tracks for mas s transi t
transportation. Horizonta l curve s have a minimum radiu s o f 25 0 m, while th e
maximum longitudina l slop e i s generall y kep t a t 2% (specia l zone s hav e 3.5%) .
The typica l bo x girde r deck ha s a constan t cros s section , Fig . 3 .
108 Muller
Fig. 3 Monterre y rapid transit system, typica l cross-sectio n
A tota l o f 650 0 segment s wer e fabricate d in a yard where 2 0 long bed s

were equippe d wit h 26 moulds fo r typica l segment s an d 1 4 moulds fo r pie r
segments. Transvers e pretensioning wa s use d i n th e dec k slab . Segment s wer e
placed i n th e viaduc t with 8 assembl y trusse s restin g temporaril y on pie r brackets.
External prestres s tendons assemble d th e segment s o f eac h spa n int o a monolit h
after matchin g th e segment s a t the col d multiple-keye d joints .
The assembl y schematic s appea r in Fig . 4 . Th e sectio n o f a typical spa n
was generall y complete d i n tw o t o thre e working days .
>NSFER
4 2 m C O N T I N U O U S S P AN
Fig. 4 Monterre y rapid transit system, assembl y schematic s

Noteworthy i s th e concep t fo r th e desig n and constructio n o f th e 1 8

elevated station s regularl y located alon g th e lin e a t 110 0 m intervals . Specia l
post-tensioned bracket s and floor beam s enabl e th e statio n platform s t o b e
supported b y th e sam e typica l dec k cross-section , Fig . 5 . Span s ar e reduced an d
the dec k i s mad e continuou s ove r th e lengt h o f th e statio n t o kee p th e stresse s i n
the bo x girde r within allowabl e values . Th e resul t i s aestheticall y pleasin g
whether see n fro m th e outsid e o r from platfor m level.
The entir e project , including desig n and construction , was complete d i n
two an d a hal f years; constructio n proper was complete d i n less than tw o years.
The firs t segmen t wa s cas t in Octobe r 1988 , while the las t o f th e 61 9 spans wa s
assembled i n Jul y 1990.
2.3 Bangko k Expresswa y Syste m

When completed , thi s projec t will constitute, probabl y fo r som e time , th e world' s
largest us e o f precas t segmental constructio n for a n urban highway bridg e
project, Fig . 6 . Face d wit h overwhelmin g traffi c problems , th e cit y o f Bangko k
has undertake n th e constructio n o f a n elevate d expresswa y system , th e secon d
FIRST STAG E EXPRESSWA Y
SECOND STAC E EXPRESSWA Y
TrCRQ S T A C E E X P R E S S W A Y
Fig. 6 Bangko k Expresswa y system ,

key pla n
110 Muller
stage o f whic h i s presentl y i n ful l constructio n (SE S project). I t comprise s a total

of 157 5 spans wit h a n averag e length o f 4 2 m built usin g precast segmenta l
construction technique s representin g a length o f 66,00 0 m an d a dec k are a o f
770,000 m . 2
The bo x girde r deck i s o f a constan t 2.4 0 m dept h an d has typicall y tw o

different width s o f 12.2 0 m an d 15.6 0 m. I n the transitio n areas , tw o bo x girders
are transversel y connecte d wit h a cast-in-plac e closur e joint. Th e to p sla b i s
transversely post-tensioned , whil e span s ar e typically simpl y supporte d t o allo w
for differentia l settlement s consequen t t o poo r an d variable soi l conditions .
Spans ar e connected longitudinall y wit h a roadway closur e i n group s o f
four t o reduc e th e numbe r o f expansio n joints . Eac h spa n comprise s 3.4 0 m long
precast segments , assemble d o n col d joints wit h multiple key s b y externa l tendon s
deviated alon g thei r trajector y and anchore d in th e en d pie r segments .
A tota l o f 20,50 0 superstructur e segments mus t b e produce d (14,50 0 fo r
the initia l par t o f th e projec t and 6,00 0 for th e incrementa l portio n thereof ) a t a n
average o f 75 0 segments pe r month thu s completin g th e initia l stag e i n tw o years.
Short-line castin g wa s selecte d becaus e o f th e variabl e geometr y
requirements an d th e limite d amoun t o f spac e availabl e fo r th e precastin g yard
(200,000 m ) . A s show n o n th e segmen t precastin g yard map, Fig. 7 , th e yard
2
was divide d into fou r productio n zones ; stee l cage s were preassemble d i n a
central locatio n whil e concret e was produce d i n tw o batc h plants.
There ar e 46 castin g machine s (3 2 for typica l segment s an d 1 4 for pie r
segments). Productio n rates are one typica l segmen t pe r da y in eac h machin e
and tw o pie r segment s pe r week i n eac h machine. Segmen t productio n starte d in
February 199 1 with the firs t tw o machines . I n August 1991 , production reache d
the assigne d goa l o f 75 0 segments pe r month whic h wa s exceede d i n Novembe r
OEMMMggggggg
rarnajwggggggg
iffliiimro*2sB!!!s
3 L_
-f-U 1 *" 1 . 1
1r m
. REINFORCIN G STEE L STORAGE .
. POS T TENSIONIN G STORAGE .
ji 1 1 i i 1 : A G G R E G A T E , CEMENT . TESTIN G LA B
i
RESTAURANT ( 2 0 0 PERSONS )
II
Mi. ; j ii PALLET STORAG E (CEMENT , HARDWARE , J A C K S ) .
-.4. —jj
. VEHICL E AN D MACHIN E MAINTENANC E SHOPS .
PRODUCTION OFFICES . SURVEYORS . QUALIT Y CONTRO L
• INFIRMER Y / FIRS T AI D
: FUE L ISLAND " AND STORAGE-FACILITY .
: ADMINISTRATIO N OFFICES . ENGINEERIN G OFFICE S
Fig. 7 Bangko k Expresswa y system. Layou t of precastin g yard

1991 with 90 0 segments .

A projec t of suc h magnitud e presente d problem s o f unusua l interest . Fo r
example, superstructur e shop drawing s ha d t o be produce d in number s an d a t a
rate heretofor e unknown . Three-dimensiona l CAD D model s wer e prepare d and
followed by almos t fully-computerise d drawing s givin g all geometr y dat a of eac h
span an d al l relevan t detail s o f eac h segmen t (dimensions , geometr y control
points, reinforcin g steel , post-tensionin g details , embedde d items , etc.) .
Construction o f substructur e works an d assembl y o f th e segmenta l
concrete superstructur e wer e divide d int o fou r section s wit h independen t genera l
contractors fo r eac h contract .
Assembly o f al l segmenta l superstructur e calls o n th e span-by-spa n
technique, wit h assembl y trusse s working underneat h the deck , o r above th e
deck, an d segmen t handlin g fro m th e to p i n both case s t o reduc e traffic
restrictions a t stree t level .
2.4 A n innovativ e desig n o f composite constructio n (stee l an d concrete ) for

medium spans : th e Roiz e Bridg e
An attemp t was mad e t o optimis e th e us e o f structura l stee l an d high-strengt h
concrete int o a new structura l schem e combinin g minimu m weight , maximu m
rigidity an d lower cost . A smal l bridg e was selecte d fo r thi s purpos e near
Grenoble, Franc e o n a grade separation of th e A4 9 tollway, Fig . 8 . Basically , th e
deck structur e comprises th e followin g components :
(a) Precas t high-strength concret e slabs (14 0 mm thick ) carryin g the highwa y
traffic an d constitutin g th e to p flang e o f th e structura l system. Desig n concrete
strength was 8 0 MPa (o n cylinder s a t 2 8 days). However , a n actua l strength of
CROSS SECTIO N
Fig. 8 Th e Roiz e bridge . Dimension s

112 Muller
100 MP a was reache d (with 4 5 kg o f silic a fume , 45 0 kg o f cemen t pe r m an d a

3
water/cement rati o of 0.2 9 made possibl e wit h plasticizer) . Concret e slabs are
longitudinally pretensione d an d transversely post-tensioned .
(b) A serie s o f stee l floor beam s a t 4 m interval s carryin g the concret e slabs
and made composit e therewith .
(c) A spac e stee l trus s including a singl e botto m flang e pip e connecte d t o th e
floor beam s b y a serie s o f diagonal s locate d in tw o incline d plane s wit h a singl e
Warren configuratio n i n eac h plane. Thi s spac e syste m i s therefor e designe d a s
the simples t assembl y o f individua l triangles where al l loads applie d ar e basically
resisted b y axia l loads i n al l structura l member s with minimu m bendin g stresses .
(d) Th e structura l capacity of th e trus s is complemente d b y external
post-tensioned tendon s deviate d alon g thei r trajectory i n muc h the sam e fashio n
as i n a precas t concrete segmental bo x girde r design .
Design-wise, th e structur e is very light (approximatel y half th e weight o f

an optimise d concret e box o f th e sam e dimensions) , whil e remaining rigi d under
live loads . Construction-wise , th e dec k is segmentall y fabricate d and assemble d
span by spa n o n a truss.
A comprehensiv e tes t programm e was performe d o n a full scal e loa d
model comprisin g fou r bays o f th e deck , which entirel y prove d th e adequac y of
the desig n o n th e followin g critica l points: (i ) fatigue o f thi n dec k slab ;
(ii) adequac y o f th e connection s betwee n individua l stee l member s an d dec k sla b
between stee l an d concrete ; (iii) respons e o f th e structur e to bending, shea r and
torsion.
The potentia l application s o f thi s ne w desig n ar e for repetitiv e medium -
span viaduct s o r light rigi d deck s fo r very long cable-staye d spans . Amon g
further refinements , presen t studie s includ e concret e filling o f th e botto m flang e
steel pip e an d simplificatio n o f th e sho p an d field connection s betwee n stee l
members.
3 Long-spa n girder bridges
Precast segmenta l constructio n has been use d extensivel y i n lon g spa n bridge s
over water or deep valley s fo r many years. Th e autho r was privilege d t o b e
associated with man y suc h projects , particularly in th e USA , includin g th e
Albemarle Soun d Crossin g i n Nort h Carolin a and the Wand o River Bridge near
Charleston, Sout h Carolina . Othe r outstanding application s ar e mentione d
below.
3.1 H- 3 Windward Viaduct, Hawaii, US A

This bridg e i s locate d o n th e H 3 Freewa y north of Honolul u o n th e islan d o f
Oahu. I t will connect th e norther n shores o f Kailua and Kaneohe with Pear l
Harbour acros s the mountai n range s separating the nort h and sout h coast s o f thi s
beautiful island . Twi n bridge s carr y th e tw o separat e traffic lanes o n a total
length o f 204 0 m , Fig . 9.
T O T A L L E N G T H O F GRIOG E = 2 0 1 li
4 CONTINUOU S SPAN S - 34 8
TYPICAL SPA N 8 7
PARTIAL ELEVATIO N
TUNNELS - ^ " ^ v ^ .
PLAN VIE W
Fig. 9 H- 3 Windwar d Viaduct , Hawaii . Partia l elevatio n an d pla n view.
Initially th e desig n featured a cast-in-plac e cantileve r dec k an d cast-in -

place bo x pier s restin g o n larg e precast concrete pile foundations . Th e
contractor's redesig n (prepare d by JMI ) offere d tw o majo r items o f interest : (1 )
use o f drille d shaf t foundation s enablin g a substantia l reductio n o f th e impac t o n
the protecte d environmen t (2 ) a precas t concrete superstructure generally mad e
integral with the pier s t o improv e seismi c behaviour . Whil e enablin g a faste r
construction schedule , precas t segmental constructio n als o reduce s th e amoun t o f
work t o b e performe d a t th e constructio n site . Unfortunately , externa l
prestressing wa s no t allowe d o n th e projec t and it was no t possibl e t o deriv e
maximum benefit fro m th e method .
The mos t interestin g aspec t o f th e projec t relates t o it s exceptiona l
geometry requirements . Th e twi n bridges follo w differen t trajectorie s with a very
significant longitudina l slop e (5 % average on th e overal l length) , variabl e curve s
in pla n (differen t betwee n th e tw o bridges ) an d relate d variations o f th e
transverse superelevation .
Segment prefabricatio n was naturall y handled wit h short line castin g t o
accommodate th e geometri c constraints . Th e precastin g yard was se t u p 1 5 miles
from th e constructio n sit e wit h four castin g machines . Segment s ar e trucked t o
the bridg e sit e wher e the y ar e handled wit h a specia l launchin g gantry , Figs 1 0
and 11 , which enable s simultaneou s constructio n o f th e twi n bridges i n spit e o f
the geometri c variations o f thei r respective layouts . Thi s ne w patente d launchin g
gantry has prove d t o work very successfully . Basicall y i t consist s o f th e followin g
components:
(a) A serie s o f cros s beams attache d to th e concret e deck which provid e th e

successive bearing s o f th e launchin g gantr y proper while compensatin g fo r th e
transverse superelevation .
114 Muller
Fig. 1 0 H- 3 Viaduct , Hawaii . Cross-sectio n o f twi n viaduct s
Fig. 1 1 H- 3 Viaduct , Hawaii . Launchin g gantr y schematic s
(b) Tw o longitudina l paralle l stee l trusse s restin g o n a t least tw o o f th e cros s

beams. On e trus s i s locate d abov e th e downstrea m we b o f th e downstrea m
bridge whil e th e othe r i s abov e th e upstrea m web o f th e upstrea m bridge . Suc h
configuration clear s th e wa y fo r th e segment s t o b e place d ont o eithe r bridg e
without movin g th e launchin g gantr y crosswise .
(c) A porta l cran e travel s th e stee l truss , picking up th e segment s a t th e rea r
from th e finishe d bridg e t o plac e the m int o thei r fina l positio n in balance d
cantilever; thi s porta l cran e ha s on e fixe d le g (excep t fo r a rotationa l capacit y
around a vertical axis ) abov e th e upstrea m bridg e whil e th e othe r le g wit h
adjustable vertical height ha s th e necessar y degrees o f freedom t o adjus t to th e

bridge geometry .
(d) A ver y simple temporar y attachment of th e porta l crane t o th e concrete
bridge enable s th e tw o longitudina l trusse s t o b e launche d separatel y to th e nex t
pier. I n fact, because constructio n proceeds downgrad e i n th e directio n o f th e
longitudinal slope , advancemen t o f th e launchin g gantr y consists rathe r i n safel y
holding back the trus s from slidin g forward spontaneously .
3.2 Locatio n of expansion joints i n long-spa n cantileve r bridge s

This i s a n important design aspec t which ha s long-term consequence s o n th e
performance an d rideability o f long-spa n concret e bridges.
Historically, th e firs t prestresse d concret e cantilever bridges wer e provide d
with a n expansio n joint a t the centr e of eac h spa n with a shea r transfer
connection allowin g free longitudina l movement s whil e maintaining th e sam e
vertical displacement s o f th e tw o adjacen t cantilever arm tips. Experienc e
showed thes e structure s to b e ver y sensitive t o long-ter m material properties
(steel relaxatio n and concret e creep); larg e permanent vertical deflection s
occurred a t the centr e of eac h spa n accompanie d by significan t angl e breaks
objectionable fro m th e twofol d aspec t o f aesthetic s an d continuit y o f th e roadway
or rai l profile .
The autho r was privilege d t o desig n the firs t precas t segmental cantileve r
bridge with continuou s dec k (Choisy-le-Roi) . Th e individua l cantilevere d arms
are connecte d a t eac h spa n centrelin e with continuity tendon s ensurin g th e
performance o f a monolithi c structure . Th e proble m o f long an d short-term
deflections an d angl e break s is provide d with a final solution .
However, othe r aspects o f th e effec t o f concret e creep and thermal
gradients late r appeared in othe r such continuou s cantileve r bridges. Significan t
tensile stresse s develo p o n th e botto m flang e nea r the spa n centr e due t o th e
combined effec t o f momen t redistributio n an d temperature gradients. Desig n
assumptions ha d t o be revise d t o properl y account for thes e factor s which wer e
originally ignore d o r under-estimated. Today , most long-spa n cantileve r bridges
have continuou s decks , an d realistic design assumptions no w ensur e their
excellent performance.
There remains , however, th e proble m o f expansio n o f long deck s (sa y
above 300-40 0 m long) . Wher e should th e expansio n joints b e locate d alon g th e
span when needed? Th e answe r appear s clearly while considerin g th e grap h in
Fig. 1 2 which show s the various deflectio n curve s under live load fo r differen t
locations o f th e hing e alon g on e spa n o f th e Olero n Viaduc t in France . Whe n
the hing e i s locate d a t the poin t o f contraflexur e of dea d load moment s (clos e t o
the quarte r spa n point) , ther e is n o differenc e betwee n th e continuou s an d
hinged structur e under dea d load, while the liv e load angl e brea k and vertical
deflections ar e significantly reduced . Consequen t t o thes e studies , i t was decide d
to generalis e th e us e o f a quarte r spa n in lieu of a centr e span hinge . Lon g
bridges suc h a s St Clou d i n Paris , Sallingsun d i n Denmark , Columbia Rive r an d
Zilwaukee Bridge s i n th e US A wer e designe d accordingly . Mor e recently, the H -
3 Viaduc t i n Hawai i follow s th e sam e principle . However , i t mus t be recognise d
116 Muller
PIERS
C0*f
\
"HINGE"
5V
-I
Fig. 1 2 Olero n Viaduct . Deflectio n curve s under live loa d
that th e desire d performanc e for rigidit y an d geometry is attaine d a t the expens e

of constructio n complexity. I n particular, i t has neve r been eas y t o carr y th e
cantilever constructio n process throug h the hing e segmen t whic h need s t o b e
temporarily blocked; moreover , the layou t o f longitudina l prestres s in th e hinge d
span i s neve r a model o f simplicity . Th e newe r concep t propose d belo w shoul d
fully answe r thi s nee d fo r improvement .
According t o th e schem e outline d i n Fig . 13 , the expansio n joint i s move d
back t o th e spa n centre , where a n activ e connectio n i s provide d betwee n th e tw o
adjacent cantileve r arm s for th e purpos e o f transferrin g at al l time s (long-ter m
under dea d load , an d short-ter m under live load), a shea r force an d a moment i n
an activ e fashion; i n othe r words, th e magnitud e o f shea r load an d momen t
across th e joint ma y be monitore d a t the designer' s reques t t o restor e all th e
geometric an d mechanical properties of a continuou s dec k acros s the expansio n
joint.
DISPLACEMENT MONITORIN G
Fig. 1 3 Expansio n joint with active connectio n

Studies mad e o n a bridge i n th e desig n stag e i n Franc e ar e summarised i n

Figs. 1 4 and 15 . Th e conventiona l centr e hinge wit h shea r transfer only
exemplifies th e usua l larg e deflections an d angl e break s while the activ e
connection wit h shea r and moment transfe r shows th e proble m o f dec k geometr y
to b e unde r full control . A continuou s membe r acros s the joint rest s o n tw o
diaphragms both side s therefro m and may be adjuste d before th e bridg e i s
opened t o traffi c t o totall y compensat e al l subsequen t effect s o f stee l relaxatio n
and concret e creep.
In fact, thi s i s onl y a first ste p an d on e ma y easily imagin e th e extensio n o f
the foregoin g principle s int o a fully monitored syste m whic h coul d contro l the
structure geometr y under al l exterio r loads (traffic , wind, earthquakes , etc) at all
times. Thi s i s a n attemp t at using artificia l intelligence i n civi l engineerin g
structures.
Fig. 1 4 Deflectio n wit h conventiona l hing e
Fig. 1 5 Deflectio n wit h activ e hing e

118 Muller
3.3 Long-spa n trusse s

Another are a o f interes t i n long-spa n concret e (o r composite stee l an d concrete)
bridges lie s in th e potentia l o f trusse d structures.
Cantilever trusse s wer e abundantl y use d a t the tur n of th e centur y for
many outstandin g stee l structures ; in recen t decades , however , othe r type s o f
designs suc h a s orthotropi c steel bo x girder s became mor e popular , particularly
outside th e USA .
It i s firml y believe d tha t th e associatio n o f hig h strength concret e an d
structural stee l trusse s coul d produc e excellen t an d economica l design s in th e
field o f lon g spa n bridges .
To demonstrat e th e potentia l o f thi s approach , a preliminar y desig n was
prepared fo r th e crossin g o f a dee p valle y i n Sout h Easter n France, wit h fre e
spans o f 46 7 m. Figs . 1 6 and 1 7 show the elevatio n an d typica l cross-sectio n o f
the project.
The 21- m wide dec k i s mad e o f a high-strength prestresse d concret e sla b
resting o n thre e vertical stee l trusses . Th e botto m chor d is als o o f high strength
concrete o n eithe r sid e o f th e pier s ove r on e thir d of th e spa n length , whil e th e
middle thir d is o f stee l onl y t o reduc e th e actua l weight o f th e structure . Dec k
construction will be b y balanced cantileve r i n increment s o f 1 5 m (distanc e
between node s o n to p an d bottom flanges) .
Diagonals ar e first place d an d connecte d t o a movable for m a t top an d
bottom flang e level s t o allo w fo r concret e casting. Th e amoun t o f longitudina l
prestress ove r th e pier s t o resis t th e cantileve r moment s i s suc h tha t onl y som e
tendons ca n be place d i n th e concret e slab while th e balanc e would be externa l
below th e dec k slab .
Cost estimate s sho w this structur e type t o b e competitiv e wit h a
cable-stayed desig n having a centr e span o f 86 0 m (se e sectio n 4.4).
J_362_m
Fig. 1 6 Long-spa n truss . Genera l elevatio n

21.000
12.600
Fig. 1 7 Long-spa n truss . Typica l cross-sectio n
4 Cable-staye d bridge s
In recen t years, many interesting cable-staye d structure s have bee n designe d an d

built aroun d th e world . A fe w o f thes e wit h which th e autho r has bee n closel y
associated wil l be mentione d i n thi s overview .
4.1 Twi n bo x girde r decks wit h centre suspension

The firs t applicatio n o f th e schem e wa s fo r th e desig n and constructio n o f th e
James Rive r Bridge sout h o f Richmond , Virginia , USA , whic h opene d t o traffi c
in 1990 . Th e 45- m approac h spans ar e continued u p t o th e mai n pier s wher e th e
pylon towe r i s locate d betwee n th e typica l pier s o f th e twi n box girders.
In th e staye d spans , th e transfe r of load s betwee n th e bo x girder s an d th e
centre suspensio n is achieve d b y a serie s o f inside strut s (i n eac h box girder ) and
precast concret e delta fram e unit s betwee n th e girder s working togethe r with th e
transversely prestresse d centr e closure dec k slab .
An improve d schem e wa s late r developed a s show n in Fig . 18 . Th e sta y
load i s transferre d in equa l part s to th e to p an d bottom slab s t o obtai n a more
direct distributio n o f th e hig h concentrate d load t o th e entir e dec k cross-section .
The precas t concret e delt a fram e which i s ver y heavily reinforce d would b e
replaced by a prefabricate d steel unit .
120 Muller
Fig. 1 8 Jame s River bridge. Stee l delt a fram e - partial longitudinal sectio n
4.2 Streamline d bo x girde r deck with twin suspension

Neches Rive r Bridge , near Port Arthur , Texas, USA wa s opene d t o traffi c i n lat e
1990, Fig . 19 , and afford s a n interesting exampl e o f a fully precast structure
incorporating th e followin g noteworth y features :
(a) streamline d precas t box girde r deck. Dra g coefficien t fo r latera l wind is
only 0.7 0 as oppose d t o 1.3 4 for a typica l box girde r deck with overhang s an d
tapered webs (lik e Brotonn e an d Sunshin e Skywa y Bridges);
(b) tw o outsid e plane s o f vertical stays anchore d in th e deck' s oute r face,
resting o n precas t segmental pylons ;
(c) activ e deviatio n saddle s locate d i n th e inne r void o f th e boxe d pylo n
section t o allo w th e continuou s stay s acros s the pylo n to b e easil y monitore d an d
readjusted a s require d t o th e effec t o f concret e creep.
Fig. 1 9 Neche s Rive r Bridge. Genera l elevatio n

4.3 Iser e Rive r Bridge o n A49 Tollway betwee n Grenobl e an d Valence, Franc e
This importan t cable-staye d structur e was opene d t o traffi c in Decembe r 199 1 to
carry th e ne w 4-lan e tollwa y ove r th e dee p valle y o f th e Iser e River. A singl e
pylon was locate d alon g th e righ t bank o f th e rive r while avoidin g th e rugge d an d
somewhat unstabl e lef t bank . Tw o shor t piers ar e added i n fron t o f th e en d
abutments, Fig . 20 , to bes t contro l the roadwa y profile unde r traffic .
Another intermediat e pie r i s adde d i n th e righ t bank mainspa n t o reduc e
deck bendin g unde r live load (th e Frenc h code i s rathe r aggressiv e i n thi s
respect) whil e providin g additiona l suppor t t o th e pylo n deck syste m durin g
construction unde r variabl e wind loading .
The dec k cross-section , Fig . 21 , provides th e ultimat e streamline d shap e
304 m
R.B.
GRENOBLE
Fig. 2 0 Iser e Rive r bridge. Genera l elevatio n
P A R T I A L L O N G I T U D I N A L S E C T I ON
Fig. 2 1 Iser e Rive r bridge. Dec k cross-sectio n an d partia l longitudinal sectio n

122 Muller
formed b y tw o triangle s abuttin g o n a centra l longitudinal web . Stay s ar e place d

at 8- m interval s an d ar e located i n pair s o n eithe r sid e o f th e centr e web wit h
anchor head s locate d a t dec k soffi t level .
The reinforce d concret e pylon, Fig . 22 , comprises a boxed shaf t belo w th e
deck extendin g thereabov e t o a concret e saddle o n whic h al l continuou s stay s are
deviated. Th e crucifor m section varie s continuousl y betwee n saddl e an d dec k
levels t o follo w th e stres s flow du e t o weight, longitudina l liv e load bendin g an d
transverse wind bending .
The dec k i s a cast-in-plac e cantileve r constructio n with the centr e web cas t
first t o allo w installatio n o f th e stays , followed by th e tw o triangle s whic h ar e cast
on suspende d forms . Architectura l details hav e bee n incorporate d into th e
design shape s o f th e pylo n to improv e it s appearanc e at limited additiona l
expense.
4.4 Ver y long-spa n cable-staye d bridge s wit h stee l space-trusse d dec k

The spac e trus s concep t develope d fo r th e constructio n o f th e Roiz e Bridg e
(section 2.4 ) could b e remarkabl y extrapolated to produc e a n efficien t dec k fo r
long cable-staye d structures.
A tentativ e applicatio n wa s studie d fo r th e Ebro n Crossing with a mai n
span o f 86 0 m, Fig . 23 . Th e dec k typica l cross-sectio n i s show n in Fig . 24 . Ther e
are tw o plane s o f stay s o n th e edg e o f th e dec k with stays locate d a t 2 0 m
intervals. Surprisingly , th e result s o f a scal e wind model tes t sho w that th e trus s
TRANSVERSE ELEVATIO N LONGITUDINA L ELEVATIO N
Fig. 2 2 Iser e Rive r bridge. Pylo n dimension s

Fig. 2 3 Ebro n Crossing. Elevatio n
10.250 750 75 0 10.25 0
Stays - Stays
Fig. 2 4 Ebro n Crossing. Typica l cross-sectio n
has a dra g factor less favourable tha n a streamlined box with continuou s surface s
To minimis e th e wind effects o n th e deck , i t will be necessar y to instal l buffer s
outside th e dec k edge s t o guid e th e airflo w awa y from th e member s o f th e spac e
truss. Cos t estimate s sho w th e 86 0 m cable-staye d schem e t o be equivalen t t o
the composite , 46 7 m trus s scheme (se e sectio n 3.3).
4.5 Th e bi-stayed concep t

All bridge s wit h a clea r span longe r tha n 500 m hav e bee n buil t t o dat e a s
suspension bridges . Currently , the world's longes t spa n i s i n th e desig n an d
construction stag e i n Japan , wit h a clea r span o f almos t 200 0 m . Meanwhile , th e
longest cable-staye d bridge s ar e Annacis Bridg e in Canad a with a clea r span o f
465 m an d th e Normand y Bridge in France , wit h a clea r span o f 85 6 m.
With existin g material s (steel wit h a yield strength of 35 0 MPa and hig h
strength concret e of sa y 10 0 MPa), the maximu m span o f a conventiona l cable -
stayed dec k lies between 100 0 and 150 0 m. Th e critica l factor is th e magnitud e
of compressiv e stresse s i n th e dec k aroun d the pylo n consequen t t o th e
124 Muller
cumulation o f th e horizonta l componen t o f th e sta y loads. Usin g a radiating sta y

configuration reduce s thi s loa d t o hal f th e valu e reache d with a harp-shape d
layout (al l stay s parallel).
To increas e th e clea r spa n beyon d th e limi t whic h correspond s t o th e
saturation o f th e dec k sectio n wit h axia l compressio n stresses , i t i s necessar y t o
find a stati c scheme enablin g an y additiona l load s t o be supporte d withou t
increasing th e dec k stresses . Th e principle , Fig . 25(a) , i s t o suspen d th e mai n
span dec k load s b y tw o differen t familie s o f stay s (whenc e th e expressio n bi -
stayed bridge) ; travelling fro m th e pylo n towards th e spa n centr e line, on e will
find:
on a length a l (convenientl y th e lengt h o f th e sid e span) , conventiona l

stays (SI ) symmetrica l with regard t o th e pylo n centreline ar e self-anchored int o
the dec k an d creat e a maximu m axia l load Nl ;
on a lengt h a2 , another family o f stay s (S2 ) which ar e now symmetrica l

with regar d to th e spa n centrelin e (an d not th e pylo n centrelin e a s the y appea r in
Fig. 25(b)) , t o balanc e th e successiv e pair s of uni t load s W in developin g a tensil e
axial loa d T 2 in th e centr e area o f th e mai n spa n rathe r tha n a compressiv e axia l
load nea r the pylons . T o ensur e equilibrium , th e stay s must b e deviate d abov e
the pylon s an d anchore d away from th e dec k in a n eart h abutment simila r to th e
anchor block s o f th e conventiona l suspensio n bridges .
The diagra m of dec k axia l load alon g th e spa n which appear s in Fig.
25(c), show s tha t the tensil e axia l load N 2 will add to th e compressiv e axia l load
Nl t o balanc e th e tota l horizontal componen t o f sta y loads fo r th e ful l hal f spa n
a = a l + a2 . Fo r the sam e materia l properties, it i s therefor e possible t o
theoretically increas e the maximu m clear span o f a cable-staye d bridge i n th e
proportion o fV"2 = 1.4 1 an d therefor e a 150 0 to 180 0 m spa n shoul d becom e
feasible. However , on e drawbac k of th e syste m lie s in th e fac t tha t heavy
permanent tensil e stresse s exis t i n th e dec k a t mid-span .
There i s a way t o compensat e fo r thes e tensil e stresse s while
simultaneously increasin g th e maximu m span very significantly. Longitudina l
post-tensioning tendon s wil l be installe d i n th e dec k t o creat e a load P whic h
compensates fo r th e tota l axia l load create d by dea d an d live loads, Fig . 25(d) .
When th e dec k i s full y loaded, ther e is a neutral state o f axia l stress, an d
when th e dec k i s unloade d (dea d load s only ) th e axia l load i s reduce d t o a
compression loa d N 2 = P.S/G+ S ( G is dea d loa d an d S Live load). Therefor e N
= N l + P = N l ( 2 + G/S) .
For a long spa n highwa y bridge , G/ S i s equa l t o abou t 3 an d N = V"5N1.
Consequently, th e limitin g spa n i s increase d in th e proportio n 2.2. Thi s principl e
makes a clea r span o f 250 0 t o 300 0 m no w feasible .
A remarkabl e advantage of th e propose d schem e i s tha t a full balance o f
all dea d load s ca n be attaine d in term s of dec k axia l load. I n particular, th e dec k
remains insensitiv e t o th e rapi d increase o f th e sta y weight i n very long spans .
Many practica l aspects mus t b e considere d i n th e implementatio n o f th e
new concept , suc h a s aeroelasti c stability an d constructability . T o preserv e the
advantages o f th e method , cantileve r constructio n should b e use d fo r th e entir e
main span .
A desig n wa s prepare d according to th e bi-staye d concep t fo r a 120 0 m
clear spa n carryin g a combinatio n o f si x highway traffi c lanes an d two trains . Fig .
26 show s th e elevatio n an d Fig. 2 7 the typica l cross-section . Nea r th e pylon , th e
box sectio n i s entirel y o f concret e while a composit e concrete/stee l sectio n i s use d
on a length o f 80 0 m i n th e centr e span.
Fig. 2 8 shows th e compariso n o f th e dec k axia l load durin g construction
and in th e finishe d bridge . Ther e is onl y a 15 % increas e (22,30 0 t versus 19,30 0
t) o f dec k axia l load durin g construction , which i s acceptable.
It i s interestin g t o compar e the vertical deflection response s o f a bi-staye d
bridge unde r live load with thos e o f a conventiona l suspensio n bridge . Fo r the
example mentione d above , subjecte d t o a highway loadin g o f 3. 4 t/m an d a train
loading o f 8. 4 t/m, th e tw o deflectio n curve s ar e shown i n Fig . 29 . Th e
superiority o f th e bi-staye d bridg e is evident .
126 Muller
Expansion joint
1
Earth a n c h o r a g e Earth anchorag e
1
1
750 m 400 m 750 m
Self Anchored Stays Earth Anchored Stays Self Anchored Stays
550 m i 800 m |1 550 m
Concrete Deck T ~ Composite Deck ' Concrete Deck
350 m 1 1 200 m | 350 m
Fig. 2 6 Bi-staye d bridge . Elevatio n
Fig. 2 7 Bi-staye d bridg e cross-sectio n
| 350 m j 330m 540 m._ j 330 m _

Deck prestres s
ELEVATION
' 22.300 tons 19.300 tons
BRIDGE I N CONSTRUCTIO N BRIDGE I N OPERATIO N

DCCK A X I A L LOA D A T CLOSUft C DCCK A X I A L L O A D UNDF.lt DC-A O L O A D A.M O POS T TCNSIONIN G
Fig. 2 8 Bi-staye d bridge . Compariso n o f deck axia l loa d

8 . 4 Tons/ m
3 . 4 Tons/ m 2 3 2 Ton s (truck )
Fig. 2 9 Bi-staye d an d suspensio n bridges . Compariso n of deflection s
To deriv e th e ful l benefit o f th e bi-stayed concept , ther e are new feature s

which nee d t o b e seriousl y considered :
(a) var y the dec k desig n alon g th e spa n by associatin g th e material s in their
most efficien t use;
(b) i n particular , use o f a composit e sla b desig n comprisin g a stiffene d stee l
plate (use d durin g construction) , subsequently covere d with high-strengt h
concrete fo r th e structur e in operation ;
(c) us e highe r allowabl e stresse s i n th e stay s beyond thos e accepte d by mos t
current code s whic h limi t th e maximu m stress t o 45 % of th e ultimat e
strength without scientifi c basis ;
(d) us e compac t stays t o reduc e to th e absolut e minimu m th e cross-sectiona l
area expose d t o wind, becaus e i n very long-span cable-staye d bridges, wind
on th e stay s is th e majo r component o f th e tota l wind load.
5 Conclusion s an d future prospects
Precast segmenta l concret e has ha d a n interesting history , an d stil l ha s a bright

future. Th e metho d prove s particularl y suitable fo r building highway an d mass
transit structure s in difficul t urba n environments.
Cable-stayed structure s have increase d in popularit y worldwide i n recent
years. I n th e rang e of span s wher e bo x girders can be use d (30 0 m fo r solid
webs an d a t least 50 0 m fo r ope n web s with stee l diagonals) , cable-staye d bridge s
will not replac e the othe r designs unti l drasti c changes tak e place both i n desig n
assumptions an d constructio n requirements to reduc e the weight an d cos t o f
stays, whic h represen t today a major component o f th e constructio n expenditure.
At th e othe r end o f th e spectrum , the ne w bi-staye d bridge s i n a field
heretofore reserve d for suspensio n bridges , enabl e mor e rigid an d more stabl e
structures t o b e buil t a t a lower cost .
Recent development s i n composit e structure s using the spac e truss
concept loo k promisin g an d shoul d ope n a new fiel d o f application s eithe r i n
medium spa n lon g viaducts o r in long span cabl e staye d bridges.
12 I T A L I A N J O I N T - F R E E
BRIDGE DECK S
F. MARTINE Z Y CABRER A
Department of Structural Engineering ,
Polytechnic of Milan, Italy
Abstract
This paper discusses the state of art, in Italy, of joint-
free bridge decks.
After some considerations about the design and the
construction systems, the paper analyzes some of the most
important italian bridges.
1 Introductio n
In the last ten years in Italy a shift has occurred toward

the adoption of a continuous beam scheme for motorway
viaducts, using both constant and variable cross sections.
This shift has been favored by the remarkable evolution
of the bearings. Moreover, bridges without joints present
several advantages in the seismic zones.
Some disadvantages of the continuous beam can be
summarized as follows;
- stresses induced by prestressing, temperature varia
tions and settlement of foundations. With an appro
priate monitoring process, however, it is possible to
know the load variations on the bearings due to the
above mentioned actions. The present technology allows
one to counteract possible settlements of the piers with
variable settings of the bearings (autoleveling
bearings);
- presence of prestressing cables in the deck,which weaken
the deck in sections with negative moments. The cables
are exposed to corrosion processes if the deck becomes
cracked or if the impermeabilization process is not
correctly executed;
- high prestress losses due to friction for long cables.
If the length of the cable is reduced, the number of
anchorages is increased and this require the solution of
many local static problems.
The type of continuous viaducts built in Italy in the
last years may be divided into three classes:
Class A: Viaducts with solid or hollow cross section (in

general a single cell) built in place, span by

130 Martinez y Cabrera
span. Mobile launching steel girders are

employed. The girder can have a costant or
variable cross section (Figs, 1.1, 1 . 2 ) .
Class B: Viaduct with precast beams mounted on adjacent

piers and connected through post-tensioned
systems. The cross section may be constant or
variable. These prefabricated elements of large
dimensions are mounted with launching steel
girders and progressively connected to the
preceding span.
The author has designed some viaducts dividing longitudi
nally the box beams in two halves. Each half has been
precast separately, and after the positioning it has been
connected in the trasversal and longitudinal direction.
c a s t in ptace
Fig. 1.2
Class C: Cantilever construction from pier with box-

girder section (in general a single cell).
The segments may be precast or cast in place,
and with constant or variable section (Fig. 1.3)
Fig. 1.3
2 Th e desig n
It is well known that the design and the calculations of

continuous beam viaducts in prestressed reinforced
concrete have many complications due to the hyperstatic
bending moments induced by prestensioning and due to the
progressive variation of the static scheme.
Italian joint-free bridge decks 13 1
It is moreover necessary to study in depth the choice

of the spans. For very long viaducts it can often be
convenient to choose an average repeated span and to adopt
some bigger spans in special situations. In this case it
is possible to adopt special segments of beam with
variable height.
For example for the Restello Viaduct on the motorway
Vittorio Veneto - Pian di Vedoia (that will be shown
later) spans of 60 m length average and special spans of
100 m length have bean designed. These special spans are
built with segments of variable height.
The design problems can be summarized in this way
*e
- relationship between outside and inner span: y r = 0,75
(Fig. 2.1)
It is necessary to choose the outside span always bigger
than 0,55 l^. Otherwise the reaction of the bearings can
change the sign for the loads. For l < 0,55 1^ it is
e
necessary to some counterweight (Fig. 2 . 2 ) ;
Fig. 2.1
counter weigh t
Fig. 2.2
- normally the height of the cross-section can be constant

for values of the span till 60-80 m.
Over 60 m it is necessary to increase the depth of the
lower deck near the internal bearings (Fig. 2 . 3 ) .
Fig. 2.3
- generally over 60 span m span it is convenient to adopt

a variable longitudinal profile with a relationship
between the height on the pier and the height in the
middle span about 1,8 - 3,0.
The height variation should ensure that the resultant of
132 Martin ez y Cabrera
the compression stresses and the tension stresses have a

linear variation along the span so that it is possible
to have constant depth for the webs.
For bridges with variable profile we have the following
relationships: (Fig.2.4).
H
i 1 _ 1 Hm 1 1
li ~ 16 20 li ~ 10 50
Normally we adopte the following variation of the
profile.
Fig. 2.4
When we have very big spans it is convenient to have

double bearings so that it is possible to have an almost
fixed-end beam (Fig. 2.5) (Ponte di Tadasuni).
Fig.2.5
2.1 Cross-section s adopte d i n Ital y

Generally the cross-sections adopted in Italy for conti
nuous bridges are the following:
a) MONO-CELLULAR BOX-GIRDER (Fig. 2.6)
Fig. 2.6
b) BOX-GIRDER WITH PROTECTED CANTILEVER SLAB (Fig. 2.7)
Fig. 2.7
c) TWO CONNECTED BOX SECTION (Fig. 2.8)
IQI—or
Fig. 2.8
d) TWO BOX-GIRDER SECTIONS (Fig. 2.9)
Fig. 2.9
e) CROSS SECTIONS CONNECTED WITH PARTICULAR PROCEDURE
Fig. 2.10 I
]f
The single segments of these cross-sections are the
following (Fig. 2.11)
Fig. 2.11
f) CROSS-SECTIONS FOR PRECAST BEAMS (Fig.2.12)

(L = 20 - 32 m)
Fig. 2.12
g) SHEL L SECTION S ( L = 2 0 - 2 5 m) (Fig . 2.13 )

With thi s constructio n syste m w e hav e a precas t concret e
shell sectio n an d w e pu t insid e i t th e ordinar y an d pre -
stressed reinforcement .
Afterwards i t i s possibl e t o complet e th e fina l sectio n
with concret e cas t i n place .
The advantag e o f thi s syste m i s th e minimu m weigh t o f th e
precast element .
Fig. 2.1 3
2.2 Typica l sectio n (Fig . 2.14 )
Fig. 2.14
A statistica l stud y du e t o Prof . Menditt o an d Prof . Dez i
(Ancona) show s th e result s reporte d i n Fig . 2.15 .
20 4 06 08 0 10 0 12 0 14 0 16 0 18 0 20 0
Span o f th e bo x b e a m
10 2 03 04 05 06 07 08 09 0 10 0 11 0
Span o f th e bo x bea m
Fig. 2.1 5
2.3 Continuit y system s

Figures 2.16a - 2.16b show some solutions designed to
obtain the structural continuity on the piers:
solution
Fig. 2.16a
2~ solution
-.*"T
earn ~~ J singl e support
continuity's /
element 1 1
Fig. 2.16b
3 Structura l analysi s
Apart from the general problems of structural analysis it

is necessary to point out some particular problems.
3.1 Constructio n phase s

- It is necessary to examine the variations of the bending
moments during each phase of construction method.
Fig. 3.1 shows the basic difference that we have with
the different construction methods.
- It is also important to examine the live load due to the
launching steel girder during its motion and during the
bridge construction.
3.2 Profil e o f prestressin g cabl e

The profile of prestressing cable must follow the
variation of bending moments from positive to negative
values.
The losses of friction can be very big.
The layout of the cables, moreover, must agree with the
construction method.
j$ ^ 7f £ i & i 2 k2 k2 k A
Fig. 3.1
It is convenient to use short cables for the negative

bending moments. The positions of the anchorages must be
studied with great attention.
3.3 Hyperstati c action s du e to tendon s

It is necessary to know the hyperstatic bending moments
due to the actions of the tendons and the variations
compared to isostatic bending moments.
The hyperstatic moments can be calculated by means of
influence lines (Fig. 3.2)
3.4 Win d action s

It is necessary to take into consideration the wind s
f
actions during the construction phases.
3.5 Deformabilit y o f the cross-sectio n

The absence of the transversal connections between the
ribs and the limited values of the thickness of the
various decks involves some problems because the cross-
section may lose its shape.
The studies can be developed with various methods of
structural analysis:
Finite Element Method
Finite Strip METHOD
FOLDED PLATE THEORY
BEAM ELASTIC FONDATION ANALOGY
9c TDD
M G • Afx* = M* (X) •(fx = 1 ) MG =
2a N • DX 2 a N • DX
o o
Fig. 3.2
3.6 Precas t element s o f a bridg e wit h longitudina l linea r

connection
The study of these bridges is developed with a particular
procedure. In Fig. 3.3 is shown the analysis scheme, with
the mutual interactions. In Fig. 3.4 the congruence
integral equations are reported.
With Fourier's series method, it is possible to transform
this equation system in to a simple system of algebraic
equations.
Fig. 3. 3
B R I D G E S WIT H L O N G I T U D I N A L C O N N E C T I O N S :
INTEGRAL EQUATION S
+ G i n d (x , « ) [ + t i + l t i + 2 ( «+ P 'itu.i+l^
* i* i+1
LL L
J Gind <*.S> i-l.i<O * - 2 J G ( x , O t i i « ) d ( + J G

l d
d i f + i n d (x.{ ) t +
i l f i + 2 « )d * =
o o o
= j G d (x.«[.pi i + , ( t ) + p j +; i +2 ( «] d t+ G , n d ( x ,« [-P!_ m «) + p ; ; ; i + 2 < ©i d «

o
SVILUPPO I N SERI E
P = n k p (k ) s e n ^ ; t ( 0 = ^tdO^iili
i i
Cl., J C l n d ( x , t ) « n ^ d t -2tki, i + l) J G ( x . t ) » a ^ d t d * + 2 JG i n d (x,£)^d { =
o o
F (x) = J G (x,0 sen ^ d * ; d .00 = J G (|,*)) sen ^ d d£
o o
F
' <> = J G
x
i n d (x,*) sen ^ d {; b^ = J G i n d sen ^ d *
o o
b(k)t< .-2.(k)t^. . b(k)t! =
k
. [-P
)
P k ) , ( k ) ,+ 1 ( k )
].aW-P , ( k )
+ i+ 1 ( k )
]-b(k )
»-1 .i + li + 1,i+2
|
i, i + I + J I + I L 1 +
l J
i-l, ii P
+ l , i+ 2 J
Fig. 3.4
3.7 Limi t state s

In each phase it is important to take into account:
- serviceability limit state
- cracking limit state
3.8 Curve d bridge s

For curved bridges all the previous problems assume diffe
rent and particular complexity.
Fig. 3.5 shows the simple case of a simple beam with
linear tendons.
3.9 Dynami c problem s

It is very important the study of dynamic characteristics
of the bridge.
In a seismic zones it is necessary to develop the
dynamic linear and non linear analysis.
3.10 Bearing s and expansion joint s

One aspect of special interest in the structural analysis,
is the dimensioning of the seats of the bearings and the
expansion joints in viaducts extending over a length o f
more than 400 m.
In a seismic zones, in Italy, antiseismic bearing
devices between the structure and the piers are studied in
order to carry out the following functions:
CABLE tz.cosianl
CEOMETRY I/, , costanl
S*ctt« 5W
X
1. FfR
z
M>1
—4
Hi
EQUIVALENT IQAOS
C * * < * f l t r a U Action s
N 'N "F
A B
Distrfculttf] A c t i o n s
Ai l*)--Fb I
t
Fig. 3. 5
Fig. 3.6 a
Scheme of the frame-type device.
Fig. 3.6 b
Scheme of the shaft-type eUstfopUstfc device.
Fig. 3.6 c
- to transmit the vertical loads due to permanent and

live actions
- to allow rotation
- to allow the excursion of the superstructure due
to creep, shrinkage and thermal effects
- to transmit the horizontal actions to a pier without
appreciable movements
- to distribute the dynamic horizontal forces (braking,
design earthquake) equally among all the piers,
without appreciable movements.
The expansion joint must allow the total excursion of the
bridge due to creep, shrinkage, and thermal effects.
Fig. 3. 6 shows a multidirectional and aseismic device.
4 Constructio n system s
The system of construction adopts advanced mechanical

equipment.
These equipments provide for all cast in place phases
and all assembly phases.
They are, really, mobile yards.
In the illustration of some bridges we will show the
systems of equipment used for the cast in place or for the
launch of beams of the bridges.
5 Monitorin g syste m an d maintenanc e
Some of the more important viaducts are provided with

special monitoring equipments.
Generally the instruments adopted are the following:
- hydraulic longitudinal and transversal level meter
- probes to measure the temperatures
- biaxial inclinometers
- bearings with dynamometer for the valuation of the
loads. The bearings can have a variable attitude.
For example they can balance the effects of the soft
ground
- transducer for the displacements
- bonded strain gage for concrete and reinforcement
- bi-axial and tri-axial accelerometers (seismic)
- traffic recorder
- transducer for meteorological events
The information is stored with event recorder
Fig. 3. 7 shows the monitoring system.
MONITORING S Y S T E M
t0 AAOUTMCM t
In in
!o
<3> Q~i^ i
KD
3ih
UGENO
jL. Sr»W» yq« fo r r e M o r c t n w u
0) T A U M If
(23 Dynami c m d w c a r (xMotifi* !
3P
Fig. 3. 7
6 Conclusion s
In Italy , i n th e las t years , a larg e numbe r o f continuou s

bridges hav e bee n built . A complet e lis t woul d b e to o lon g
for inclusio n i n thi s paper .
For thi s reaso n I am describin g jus t a fe w o f them .
- GRANUD A AN D CASELL E VIADUCT S

(MOTORWAY CARNI A TARVISIO )
- RESTELL O VIADUC T
(MOTORWAY VITT0RI 0 VENET O - PIAN D I VED0IA )
7 Reference s
AISCAT - C.N.R . (1990 ) Norm e d i manutenzion e pe r pont i

stradali.
Campo, G . Corona , G . an d Mancini , G . (1990 ) Th e Torin o
Frejus highwa y bridges . Developmen t i n Shor t an d Mediu m
Span Bridg e Engineering .
Capra, R . an d Martine z y Cabrera , F . (1993 ) I I Pont e d i
Tadasuni. Giornat e A.I.e.A.P. , Pis a
Dezi, L . Menditto , G . an d Rosignoli , M . (1982 ) I n tem a d i
predimensionamento de i pont i a cassone . L'Industri a
Italiana de l Cemento , n ° 6 .
Ghinassi, G . an d Grillo , F . (1993 ) Analis i sismic a d i u n
viadotto i n c.a.p . a travat a continu a co n tracciat o
planimetrico a doppi a curvatura . Giornat e A.I.C.A.P .
Marioni, A . Marchetti , D . an d Cozzani , G . (1991 ) L a
Costruzione de l Viadott o Restello . L'Edilizia , n ° 6 .
Marioni, A . (1991 ) Antiseismi c bearin g device s o n th e
Mortaiolo Viaducts . Thir d Worl d Congres s o n Joint ,
Sealing an d Bearin g Syste m fo r Concret e structures .
Toronto, Canada . ACI .
Martinez y Cabrera , F . Rinelli , A . an d Zuccolo , G . (1987 )
I Viadott i Granud a e Casell o pe r 1'autostrad a Carnia -
Tarvisio. L'Industri a Italian a de l Cemento . n ° 609 .
Parducci, A . (1987 ) Apparecchiatur e dissipativ e special i
per l a riduzion e dell a rispost a sismica . Giornat e
A.I.C.A.P.
Rapino, F . Caloisi , G . an d Merletti , T . (1991 ) I I Viadott o
Restello. Giornat e A.I.C.A.P .
VIADUCT "GRANUDA" - "VIADUCT "CASELLO"*
MOTORWAY CARNIA-TARVISIO (ITALY)
THE VIADUCT'S GENERAL DIMENSIONS
The Granuda viaduct is 320 m long overall and has eight

40,00 m spans.
The Casello viaduct is 420 m long overall, has eight 45,00
m spans, as well as two 30,00 m approach spans.
The structures have a constant width of 13,32 m, for a
single carriageway(Fig. 1 ) .
The resulting cross section that resulted is a three-cell
caisson with a rectangular center cell; the side cells,
incorporating the usual cantilever, are triangular, their
side walls being inclined. Such a section is advisable for
structures located in areas subject to low temperatures,
since the deck's upper slab exposure to atmospheric agents
is then transversely uniform.
The bearing and aseismic devices are those commonly known,
used for all the motorway viaducts.
THE LONGITUDINAL SEGMENTING OF THE VIADUCT DESKS
The longitudinal segmenting system used to build the decks

of the Granuda and Casello viaducts presents some
interesting new basic features, since it was extended for
the first time here to large-scale box elements.
The technology adopted fits within the framework of the
segmental precasting process, but in fact calls for the
casting of span-long base elements, assembled longitudi
nally. This process, in many ways a counterpart of the
transverse segmenting system, marks a noteworthy advance
in viaduct-construction methods, having a number of
statics and construction advantages.
THE CONSTRUCTION PHASES OF T H E DECKS
Fundamentally, the construction process broke down into

the segmenting of the caisson deck girder converting the
whole span into two halves by passing a vertical plane
through its longitudinal axis, and into the connection of
the half-caissons by two continuous linear joints. Fig. 2.
* FEDERIC I COMPAN Y - (ROME) - S.A.L.C. COMPAN Y (PADOVA )

Designer: F . Martine z y Cabrer a (Milano) - A. Rinell i (Padova) - G. Zuccol o (Padova )
The decks were built in the following phases:
a) precasting at the plant of each half-caisson over its
whole length, its steam curing, and the other phases
preliminary to prestressing to prepare it for haul;
b)hoisting by hydraulic devices, hauling on rails and
laying in position using high-capacity launch equipment;
the precast elements are put side by side in place on
special complex bases set up to adjust their position in
line and grade;
c)casting the connection for top and bottom slabs, along
the interface lines;
d)laying the continuity cables between the launched span
elements, execution of the connection casts between the
girder ends, and tensioning of some continuity cables;
e)completion of cable tensioning operations when the
successive spans had been launched;
f) finish works: curbs and paving.
This solution was advantageous in several ways:

- half-caissons poured in a few hours in a stable and
protected yard, with full control of materials and of
the curing process. This meant true longitudinal
element monolithicity, difficult to obtain using other
procedures. Thus, the longitudinal cross section is
kept integral, with no statics interruptions due to
bending and shear;
- none of the interface problems inherent to other
structural systems;
- the structure as a whole is more monolithic than in-
situ-poured caissons. These latter are necessarily
built " i n i n s t a l l m e n t s " : a first cast of the bottom,
slab, then the cast of the walls, and finally the top
slab. This means a larger number of successive
castright in the statically-engaged areas, that is, in
the nodes of the caisson cross section;
- the longitudinal segmenting involves only problems
tied to local behaviour and, as regards the overall
statics, tied to its torsional behaviour in response
to antisymmetric loading conditions;
- the effects due to deferred elastic phenomena are more
easily controlled, because of the precast structural
element longitudinal homogeneity.
The continuous construction cycle described meant,
using two outside forms, an average construction time
of three days for each half-caisson (corresponding to
3 00 mq of deck area) ; after this the launch, assembly
and joining phases these not slowing down the advance
of the bridge.
a) Static analysis
Numerical analysis was used to study the caisson, using a
finite-strip discretizing procedure, and treating it as a
plate-slab. This method can be used when the elastic and
Fig. 2
geometric parameters stay constant in the longitudinal
direction, only the amount of the load varying. (Fig. 3 ) .
b) Experimental analysis
On the basis of the above analysis it can be stated that
in longitudinal assembly systems the connection joints,
even though of considerable importance, are not in
themselves a critical factor for the structure's safety.
Nonetheless experimental investigations were undertaken to
study the joints real behaviour and their aptness to bring
about the desired cross-wise static continuity.
Full scale models were then constructed of the area of the
slab affected by the joint (Fig. 4 ) , reproducing a 1,45 m-
long strip from the caisson, with a transversal profile
similar to the central area of the caisson, including the
parts where the side walls were attached. Fig. 4 shows the
model dimension, together with some of the loading
conditions studied:
- two uniform strip loads of dimensions 10 x 145, astride
the transversal mid-span and spaced 1,20 m apart;
- two load "imprints" of dimensions 65 x 25, spaced at
1,80 m;
- one isolated load "imprint of dimensions 65 x 25,
11
placed at the center of the slab.
Fig- 3
1 ft
Start of self-launching operations
,1 I . J . ' T
ir
Auxliary bearing on the new pier
3
*4
Fino/ positioning of the transversal tracks
*Syb i i J -i i j u
Connection of the half-caisson to the front winch
T i l l !
Advance-ment and connection of the half-caisson to the rear winch
11
Jli ' 1 1 1
Jl
j3 t
Positioning of a halfcaisson in its final location
Fig. 5
M O T O R W A Y OF FREJUS (ITALY)
C L A R E A A N D R A M A T VIADUCTS*
TOTAL LENGHT C L A R E A I^qj = 600 m

R A M A T LJ.Q = 900 m
T
SINGLE SPA N 1 = 10 0 m
CROSS SECTION monocellular bo x girde r 10.5 5 m wid e an d variabl e heigh t

from 2,4 5 m to 4 , 55 m with parabolic la w
(two carriageway)
PIERS box typ e u p t o 9 2 m i n heigh t
SCHEME continuos beam o n seve n span s
TYPE segmental bridg e buil t assemblin g precas t segments b y mea n

of pos t tensione d tendon s (trasversa l segmentation).
In th e assemblin g procedur e a stee l launchin g girde r ha s
been adopted , havin g a total weight o f 280 tons .
The launchin g girde r wa s capabl e o f handlin g segment s u p
to 12 0 t in weight .
BRIDGE T O PIE R CONTAC T the contac t i s obtaine d b y mean s o f tw o line s o f thre e

bearings each , space d 3,9 0 m apart (reductio n o f 10 % in th e
positive bendin g moment) .
SPEED O F COSTRUCTION costruction o f tw o complet e hal f span i n 21 day s

(4,76 m/day) .
MATERIALS concrete 0,78 m c / m 2
ordinary reinforcemen t 110 kg/m 2
presstressed reinforcemen t 45 kg/m 2
< L^ v
<2> *
* RIZZAN I DE ECKE R
Designer: G . Coron a (Turin) - G. Mancin i (Turin ) - P . Napol i (Turin )
CLAREA VIADUC T ( Frejus motorway - Italy )
Contractor: RIZZANI DE ECKER

Designer: Ing. G. Corona
Ing. G. Mancini
Ing. P. Napoli
M O T O R W A Y VITTORIO VENETO - PIAN DI VEDOIA
RESTELLO VIADUCT
TOTAL LENGHT Lq-Qj = 210 0 m
SINGLE SPA N I = 4 8 m ; 7 2 m ; 60 m ; 10 0 m
CROSS SECTION monocellular bo x girde r 12,8 8 m wid e and , variabl e heigh t

from 3,7 0 m to 5,00 m with linea r law (tw o carriageway).
PIERS box typ e u p t o 60 m in height .
SCHEME continuos bea m
TYPE segmental bridg e buil t assemblin g larg e precas t segment s b y

mean of post-tensione d tendon s (trasversa l segmentation).
The segment s ar e o f tw o basi c type : on e segmen t wit h
variable sectio n an d 4 0 m lengh t an d anothe r typ e wit h
constant section an d variou s lengh t (3 2 m ; 4 0 m ; 6 0 m )
In th e assemblin g procedur e a stee l launchin g girde r ha s
been adopted .
The launchin g girde r wa s capabl e o f handlin g segment s u p
to 80 0 t in weight .
BRIDGE T O PIE R CONTAC T two multidirectiona l bearin g wit h longitudina l excursio n S L
= 100 0 mm and trasversa l excursion S = T 1 5 mm .

The viaduc t ha s aseismi c longitudina l device s a t th e
about men ts and transvers e devices o n th e piers .
MATERIALS concrete 0,90 mc/m 2
ordinary reinforcemen t 100 k g / m 2
presstressed reinforcemen t 50 k g / m 2
Contractor: GIC O COMPANY

Designer: Italprogett i S.r.l . - Rom a - Ing . Calois i
Engineer's Representative : S.P.E. A . - Milan - Ing . T . Merlett i
Autostrada Vittori o Venet o Pia n d i Vedoi a (Italy )
Viadotto Restell o (continou s bea m L = 2100m)
A. Spans type
Italian join t-free bridge decks 15 7
VIADOTTO R E S T E L L O - A U T O S T R A D A VITTORI O VENET O - P I AN D I V E D O I A
FASI D I VAR O - LAUNCHIN G O F 100 m LON G SPAN S
-CANTILEVER HOL D
-STARTING ADVANCIN G
I • -- - - -GIRDER ADVANCIN G
1 ' l i 1
-CANTILEVER ADVANCIN G
-CANTILEVER PULLBAC X
f « a — -LATERAL TRANSFE R
-TEMPORARILY LAYIN G
-FRONT-SIDE WINC H
n n 1 „ I J ..... W V N •DISPLACEMENT
i i_ -NEW HOL D OYE R TH E TRAC K
f y — t r . ii * - 4 - -LAYING O N
PROVISIONALS SUPPORT S
-WINCH ANCHORAG E
W V N
-CIRDER ADVANCIN G PLACIN G
-REKFORCED CAG E PLACIN G
IN TH E FORtVJORX
-POURING
-GIRDER PULLBAC X
-SACK PENDULU M PLACIN G
-HOLD O f TH E BEA M
-BEAM ADVANCIN G
-WINCH ANCHORAG E
-GIRDER ADVANCIN G
-RENFORCED CAG E PLACIN G
-PULLBACX O F TH E BEA M
-LATERAL TRANSFE R GIRDE R
-PLACING O F TH E BEA M
2? -TRACK ADVANCIN G
- f ORTWOR K HOL D
W V N
-WINCH ANCHORAG E
-GIRDER ADVANCIN G
-FRONT-SIOE TRAC K
ADVANCING AN D LAYIN G
ON TH E PIE R
W V N -BACK PENDULU M HOL D

-GIRDER ADVANCIN G
-REINFORCED CAG E PLACIN G
-POURING
60 m girde r launchin g
View o f th e fron t rai l bea m Launchin g o f th e to p

pier girde r
Operator a t contro l pane l

13 TH E U K DEPARTMEN T O F
T R A N S P O R T VIE W O N
CONTINUITY/INTEGRAL
BRIDGES
G.P. P A T E R S O N
Department of Transport, London, U K
Abstract
The Department of Transport will shortly publish a new
Departmental Standard and Advice Note on "Design for
Durability", whose requirements will apply to all new
bridges built for the Department. The need to make new
bridges continuous over intermediate piers, and, for short
and medium spans, fully integral at their ends, is an
important part of these new requirements. The paper
describes the development of the documents and gives details
of the requirements for continuous and integral bridges,
together with some explanation of the thinking which lay
behind the development of these documents.
Keywords: Bridge, Durability, Standards, Continuous, Buried
structures.
1 Introductio n
The Department of Transport has been aware, at least since

the mid 1970s, that the performance of its stock of bridges
built in the post-World War II years was not as satisfactory
as had been expected. One of the main causes of this was
the deterioration caused to the ends of bridge decks and to
substructure elements below them due to the presence of full
joints at the ends of the decks, and at the ends of each
span in multi-span bridges. Accordingly, individual
submissions for approval in principle started to be vetted
with a view to eliminating fully discontinuous joints at
intermediate piers.
The "Maunsell Report", (1), published in 1989, was a
survey of a randomly-chosen set of 200 bridges, and its
findings corroborated, among other things, the perception
that the presence of full deck joints was a major source of
deterioration in the bridge stock as a whole. By the time
this report was published, however, the Department was
already preparing a first draft of an advice note on Design
for Durability, in which the issue of bridge continuity was
to be of central importance.
© Crow n copyright
162 Paterson
2 Developmen t o f th e Department' s Draf t Advice Not e
In it s earl y draf t stages , th e Advic e Not e mad e a stron g

recommendation tha t ful l o r partia l continuit y shoul d b e
used a t al l intermediat e piers , an d tha t half-joint s b e
totally avoide d whereve r possible . I t als o continue d wit h
this logi c to recommend th e use of fully-integra l dec k ends ,
that i s t o say , th e us e o f portal-typ e bridge s wit h th e
abutment stem s integra l with the ends of the deck. However ,
received goo d practic e was that such bridge types could onl y
be use d fo r quit e shor t bridg e lengths , an d therefor e
portals wer e generall y though t suitabl e onl y fo r singl e
spans o f 10 m o r less .
It gradually becam e obviou s tha t thi s perceptio n o f goo d
practice was very conservative, and that much greater bridg e
lengths coul d b e constructe d a s integra l bridges .
Experience fro m th e U.S.A . particularl y fro m th e Stat e o f
Tennessee, an d fro m Sweden , suggeste d tha t length s o f th e
order o f 100 m coul d b e achieve d withou t incurrin g problems .
Since th e distributio n o f th e number s o f ou r stoc k o f
bridges i s highly skewe d towards the shorter bridge lengths ,
the applicatio n o f a requiremen t tha t bridge s shorte r tha n
a lengt h of , say , 100 m shoul d b e designe d a s integra l
wherever possibl e woul d mea n tha t th e overwhelmin g majorit y
of al l ou r ne w bridge s woul d b e o f th e integra l type .
The limitatio n o n th e lengt h o f integra l bridge s i s
imposed b y th e respons e o f th e soi l mas s behin d th e
abutments. Lon g integra l bridge s experienc e larg e change s
in lengt h a s a resul t o f therma l effects ; i n war m periods ,
their end s push agains t th e soil mass and forc e i t to defor m
in a compressiv e manner ; give n th e hig h relativ e strengt h
and stiffnes s o f structura l materials , th e movement s o f
expanding deck s ca n b e considere d a s almos t irresistible ,
although stif f soil s ma y induc e significan t compressiv e
stresses i n th e dec k a s well .
In col d periods , th e dec k shrink s and , dependin g o n it s
nature, th e soi l wil l eithe r expan d t o wholl y fil l th e
resulting void , o r i t may rela x t o a lesse r degree , leavin g
a voi d a t th e interfac e unti l th e nex t expansiv e cycl e
closes th e gap . I t i s the cycli c natur e o f thes e movement s
which cause s concer n t o bridg e designers . Bot h dail y an d
seasonal cycle s ar e involved , an d th e numbe r o f cycle s i s
large i n the life-tim e o f a typical bridge . I t is difficul t
to forese e th e cumulativ e effect s o f man y suc h cycles ;
perhaps th e soil' s resistanc e build s u p progressivel y afte r
many cycles , implyin g a gradua l vertica l compactio n o f th e
soil mas s du e t o cumulativ e secondar y effects ; o r perhap s
some soil s hav e th e abilit y t o cop e wit h dec k movement s b y
creeping awa y fro m a n expandin g dec k an d toward s a
contracting one . Th e Departmen t ha s commissione d researc h
through TR L t o explor e thes e issue s i n mor e depth .
Meantime, existin g finding s suc h a s thos e o f Brom s an d
Ingelson ( 2 ) may b e use d fo r design .
DTP view on continuity/integral bridges 16 3
However, thi s ma y b e , experienc e abroa d wa s suggestin g
that, i n practice , n o seriou s problem s wer e arisin g i n th e
management o f suc h lon g integra l bridges . I t was decided t o
check th e Department' s bridge s data-bas e t o se e wha t wer e
the longes t integra l bridge s i n our own inventory, and what,
if any , problem s ha d bee n experience d wit h them . Give n th e
widespread acceptanc e o f th e uppe r limi t o f abou t 10 m a s
traditional goo d practice , w e wer e surprise d t o fin d tha t
the Englis h Trun k Roa d Networ k contain s a larg e numbe r o f
integral bridge s muc h longe r tha n this . Ther e ar e a t leas t
20 bridge s wit h length s i n th e 35-50 m range ; th e longes t
bridge w e hav e identifie d i s abou t 110 m long , althoug h it s
skew o f abou t 60 ° correspond s t o a righ t spa n o f approx -
imately 60m . Man y o f thes e bridge s ha d bee n buil t i n th e
1960's an d 1970' s as part of the development o f the motorwa y
system, and had been performing i n a remarkably trouble-fre e
way i n service, to the extent that the significance o f thei r
integral natur e ha d bee n largel y overlooked .
Hence i t i s no w propose d tha t th e Departmen t shoul d
require tha t al l bridge s u p t o length s o f abou t 6 0 o r 70 m
should b e mad e wit h full y integra l abutment s an d wit h ful l
or partia l continuit y a t intermediat e pier s where suitable .
The fina l for m o f th e advic e not e wil l reflec t this , an d a
short Departmenta l Standar d wil l giv e mandator y forc e t o
this an d certai n othe r requirement s relatin g t o bridg e
durability. Th e drafts of these two documents ar e now bein g
piloted throug h thei r technica l an d polic y revie w committe e
stages withi n th e Department , and , al l bein g well , bot h
should b e publishe d befor e th e en d o f 1993 . A s a result o f
these revie w procedures , some o f th e requirement s describe d
later i n thi s pape r ma y b e subjec t t o change .
3 Content s o f th e Department' s Advic e Not e
The Advic e Not e recommend s th e us e o f continuou s bridg e

decks whereve r possible , eve n i n situation s wher e
considerable relativ e settlement s ma y occu r betwee n
individual piers . Du e allowanc e shoul d b e mad e fo r
predicted relativ e movement s i n th e desig n o f th e dec k
elements. Wher e thes e effect s canno t b e catere d fo r usin g
full continuity , partia l continuit y ma y b e preferred , tha t
is, th e use of simpl y supporte d beam s with a continuous dec k
slab. However , i t i s pointed ou t that , i n fully-continuou s
decks, th e abilit y t o cop e wit h relativ e settlement s i s
enhanced b y the use of higher span/dept h ratios . Span/dept h
ratios fo r continuou s bridge s ten d t o li e i n th e rang e
between 2 0 an d 25 , bu t ratio s o f 3 0 o r eve n 3 5 ar e
achievable fo r th e sam e dec k strength , thoug h wit h a
progressive los s o f econom y a t th e constructio n stage . Th e
Advice Note points out that, if such slender decks are used ,
care need s t o b e take n t o ensur e tha t th e deck s ar e no t to o
"lively".
164 Paterson
Whereas wit h reinforce d concrete , stee l an d
steel/concrete composit e deck s ful l continuit y wil l
generally b e provided , ther e i s a proble m wit h th e us e o f
pretensioned prestresse d concret e beams , whic h d o no t
obviously len d themselve s t o continuou s applications . Ful l
continuity ca n b e achieve d b y variou s methods , bu t som e
uncertainty remain s abou t th e fina l distributio n o f dead -
load stresses , du e t o th e difficult y i n predictin g
accurately th e developmen t i n tim e o f th e effect s o f creep ,
shrinkage an d othe r losse s i n the bea m elements . Ther e ca n
be ver y significan t redistri - butio n o f dead-loa d moment s
between suppor t an d mid-spa n area s du e t o thi s effect .
Partial continuit y may be preferred, though the detailing a t
pier head s i s likel y t o b e mor e clumsy , wit h separat e set s
of bearing s fo r eac h span . Severa l alternativ e form s o f
partial continuit y ar e available ; I personall y prefe r th e
form whic h ha s bee n use d extensivel y i n German y an d wa s
pioneered i n U K b y D r A . Kumar , involvin g th e us e o f a
compressible laye r betwee n th e beam s an d th e slab , s o tha t
the deflectio n profile s o f eac h unde r loa d ca n b e entirel y
independent ove r a certai n zon e a t th e pie r position .
The Advic e Not e goe s o n t o recommen d th e us e o f fully -
integral structures , tha t is , structure s wit h ful l o r
partial continuit y a t piers , an d wit h abutment s (an d
possibly als o pier-heads ) buil t integrall y wit h th e bridg e
deck. Dec k expansio n joint s woul d b e totall y eliminate d b y
this recommendation ; i f pier-head s ar e integral , bearing s
would als o b e dispense d wit h entirely .
The use of run-on slab s at bridge abutments i s discussed .
This ma y hav e severa l advantage s whe n use d wit h integra l
bridges. Th e mos t importan t o f thes e i s t o preven t th e
occurrence o f vertica l discontinuitie s i n th e fina l roa d
surface, whic h migh t b e cause d b y progressiv e over -
compaction o f the soil locall y behind th e end of the bridge.
Such discontinuitie s ca n b e hazardou s t o traffic , an d ca n
cause bi g increase s i n th e dynami c effect s o f vehicl e axle s
on bridg e decks . Run-o n slab s ensur e tha t onl y sligh t
rotational discontinuitie s occu r a t thei r ends , alon g wit h
horizontal discontinuitie s a t thei r oute r end s only .
Another advantag e o f th e us e o f run-o n slab s i s tha t th e
soil mas s immediatel y behin d th e abutmen t ma y b e lef t onl y
partly compacted , wit h possibl e benefit s i n reducin g soi l
pressures o n abutment stems ; in general we do not favour th e
US practic e o f usin g wholl y uncompacte d fil l belo w run-o n
slabs. Conservativ e guidance on the design of reinforcemen t
for run-o n slab s i s given .
Obviously, th e magnitud e o f soi l pressure s generate d b y
long integra l bridge s ca n b e large , reachin g passiv e
pressure values . Th e us e o f slende r abutmen t stem s ma y
mitigate thes e effect s t o som e extent , sinc e th e larg e
excursions wil l the n b e mor e limite d t o th e uppe r level s o f
the soi l mass . Thi s suggest s th e us e o f "balanced-design "
for R.C. stems , that is , the concrete sectio n i s not thicke r
DTP view on continuity/integral bridges 16 5
than i s neede d t o provid e a compressiv e coupl e equa l t o th e
tensile coupl e o f th e reinforcement . Th e Advic e Not e doe s
not giv e guidanc e o n th e calculatio n o f eart h pressures ,
since i t i s assume d tha t standar d theorie s ca n b e use d t o
derive thes e fro m th e structura l movement s involved .
Designers ma y nee d t o asses s th e thrus t force s communicate d
by th e soi l t o th e deck , an d allo w fo r thi s i n buildin g u p
the overall stresse s i n the deck cross-section. Th e genera l
guidance o n thes e problem s i s that , althoug h substantia l
additional cost s ma y b e incurre d i n dealin g wit h soi l
pressures o n lon g integra l bridges , the benefit s o f reduce d
maintenance an d mor e durabl e structure s ar e likel y t o fa r
outweigh them .
Although th e genera l treatmen t o f integra l abutmen t
design i s i n term s o f th e us e o f standar d full-heigh t
abutment walls , w e ar e awar e tha t bank-sea t abutment s ar e
also ver y common , an d tha t us e o f thi s typ e i s likel y t o
increase eve n furthe r i n future . Muc h o f th e discussio n i s
common t o bot h types , o f course , bu t explici t referenc e i s
also mad e t o integra l bank-seats . Bot h U.S . an d Swedis h
practice ha s develope d specifi c way s o f dealin g wit h thi s
case, and th e forme r i s referred t o i n the Adivce Note, with
an illustratio n o f th e typica l arrangemen t used .
The Advic e Not e set s ou t th e variou s recommendation s fo r
the desig n o f integra l bridge s a s outline d above . Th e
Standard require s tha t al l bridge s shal l b e designe d a s
continuous ove r intermediat e pier s wit h full y integra l
abutments fo r bridg e length s u p t o 70 m (th e precis e figur e
may var y slightl y i n th e finall y publishe d version) .
Clearly, thi s i s no t goin g t o b e possibl e i n ever y case ;
however, we expect the number of instance s where recours e t o
"Departures fro m Standard " i s require d t o b e quit e few .
In line with the promotion o f integra l bridge structures ,
the Standar d als o require s that , wher e adequat e clearance s
exist, burie d structure s shoul d b e used . Apar t fro m th e
obvious poin t tha t burie d structure s ar e normall y integra l
structure types , (e.g . R.C . bo x culvert s an d corrugated -
steel burie d structures ) whic h ar e no t expose d t o th e
problems o f join t leakage , th e us e o f al l burie d structur e
types, includin g th e integra l types , hav e tw o furthe r
advantages:-
being relativel y remot e fro m th e highwa y runnin g

surface, problem s o f aeria l sal t spra y an d o f wate r
penetration aide d b y th e locall y hig h hydrauli c
overpressures du e t o whee l load s ar e reduce d i n
significance
being structurall y separat e fro m th e highwa y
structure, maintenanc e o f bot h i s eased .
166 Paterson
4 Conclusio n
The coverage o f the Departments' Standard an d Advice Note o n

Design fo r Durabilit y i s muc h mor e extensiv e tha n th e
treatment o f continuou s an d integra l bridg e types . Fo r
example, ther e i s a n importan t sectio n o n th e adoptio n o f
structure type s whic h minimiz e o r avoi d th e us e o f tensil e
reinforcement i n concret e structures ; a s par t o f thi s
section, th e us e o f th e arc h typ e o f structure , which i s o f
course anothe r integra l structur e form , i s advocated . Th e
Department's participatio n i n a numbe r o f initiative s t o
build arc h bridges , bot h i n concret e an d i n masonry , i s
aimed a t re-establishin g thi s bridg e typ e a s a significan t
option fo r ne w bridg e procurement .
Nevertheless, th e detaile d coverag e o f th e requirement s

for continuou s an d integra l bridges , an d th e reinforcemen t
of thes e requirement s b y us e o f a Departmenta l Standar d i s
likely to lea d to significant changes i n the overall patter n
of ou r procurement . Thi s shoul d lea d t o a progressiv e
change i n th e balanc e o f ou r bridg e stoc k toward s simpler ,
more easily-maintaine d bridges , wit h reduce d need s fo r
bearings an d expansio n joint s an d fo r th e troublesom e
facilities neede d t o servic e them .
5 Reference s
Broms, B.B . an d Ingleson , I . Eart h Pressur e agains t th e

Abutments o f a Rigi d Fram e Bridge , Geotechniqu e 21 ,
No. 1 .
Wallbank, E.J . (Apri l 1989 ) Th e Performanc e o f Concret e i n
Bridges, HMSO , Londo n
14 J O I N T L E S S S T E E L
VIADUCTS FO R RAILWA Y
E.C. HAMBL Y
Edmund Hambly Ltd, Berkhamsted, UK
G.W. OWEN S
Steel Construction Institute, Ascot, UK
Abstract
This pape r examine s th e feasibilit y o f usin g jointless , thermall y restraine d
construction fo r railwa y viaducts . Prim a facie , i t demonstrate s tha t suc h
superstructures ar e technically acceptabl e fo r bot h strengt h an d fatigu e limi t states .
It als o show s tha t th e abutment s an d pier s ca n b e designe d satisfactorily . Suc h
construction is mos t likely to be economic fo r multispan viaducts . I t is likely to show
considerable economie s i n service ; unlik e conventiona l construction , ther e ar e n o
incompatibilities o f therma l movemen t betwee n th e rails , superstructur e an d
substructure, maintenanc e costs shoul d therefore be reduced.
Key words : continuous , bridges , steel , railway , jointless, maintenanc e
1 Introductio n
Steel viaduct s ca n b e designe d u p t o an y lengt h withou t movemen t joints . Suc h

viaducts woul d b e particularl y suitabl e fo r hig h spee d railways . Th e behaviou r o f
a jointless viaduc t under thermal and live loading woul d be essentially th e same as the
behaviour o f th e continuou s welde d railwa y track . Th e absenc e o f an y expansio n
joints i n the structure obviates th e need fo r any joints or other special measure s in the
track construction . I t also eliminate s source s o f nois e at bearings an d joints durin g
passage o f trains.
It has been demonstrated in North America that concrete and steel viaducts can
function satisfactoril y withou t expansio n joint s a t either abutmen t fo r length s o f u p
to 200m , Burk e (1990) . Thes e viaduct s expan d an d contrac t during temperature
changes; thoug h i n genera l th e dail y an d annua l cycle s o f temperatur e strain are
much smalle r tha n ar e calculate d fo r cod e extrem e range s o f temperature . Th e
thermal movement s tha t d o occu r ar e accommodate d b y deformation s withi n th e
embankments behin d th e abutments. However , fo r high speed railways thes e smal l
movements stil l caus e problem s becaus e th e structure s change i n lengt h whil e th e
continuous railwa y trac k abov e doe s not. Thi s i s on e o f th e reasons wh y littl e us e
has been made of continuous girder s for railway viaducts. Stee l railwa y bridges hav e
generally bee n simpl y supported , an d hav e tende d t o be mor e noisy i n servic e tha n
concrete (thoug h composit e constructio n an d ballast d o reduc e noise considerably).
The viaduc t propose d her e ha s anchore d abutment s whic h d o no t move .

168 Hambly and Owens
Instead of expandin g o r contracting during temperature change, th e steel i s subjecte d
to locked-i n compressio n o r tension stresse s fro m en d t o end , i n th e sam e manner
as continuou s railwa y track . A t th e end s thes e stresse s ar e transferre d int o th e
abutments an d then int o anchorag e stresses i n th e ground. Th e abutments are more
massive tha n conventional abutments , however thei r anchorage requirement does not
increase wit h th e lengt h o f th e viaduct ; s o jointles s constructio n become s mor e
economic a s th e length increases .
Longitudinal tractio n and braking force s ar e transmitted along th e full lengt h
of the viaduct to the abutment anchorages. A s a result relatively slende r piers can be
used fo r all intermediate supports. Th e viaduct could be constructed at the abutments
and launche d incrementally , i n orde r to minimis e th e interferenc e wit h th e ground
below. Onc e th e superstructur e has been mad e continuous wit h th e abutments , th e
bearings use d fo r constructio n ca n b e fixed . I f provisio n ha s t o b e mad e fo r
replacement of span s in the future the abutments and bearings can be designed so that
they ca n be unfixe d whil e the replacement take s place.
A chec k ha s t o b e mad e of th e fatigu e fro m therma l stres s cycle s a t welded
joints an d at the holes for friction gri p bolts; howeve r thi s fatigu e doe s not appear to
be critica l as compare d to th e design fo r live loading .
It is foun d tha t the maximu m stres s rang e induced b y extreme hot an d col d
temperatures i s smal l compare d t o th e stres s rang e capacit y o f hig h strengt h stee l
between compressio n an d tensio n limits . I n contrast, a jointless viaduc t of concrete
appears to be less practicable because th e stress range between extrem e temperatures
would us e u p a hig h proportio n o f th e compressio n capacit y o f th e material . Th e
system i s also inherently les s suitable fo r concrete because of the much greater cross-
sectional are a o f bridge s o f comparabl e spa n fo r tha t material . Eve n allowin g fo r
differences i n modulus, th e thermal forces induced in a concrete superstructure would
be severa l time s greate r than in a steel superstructure.
The jointless stee l viaduc t demonstrated here is assumed t o have conventiona l
protective treatment , o r t o b e o f weatherin g stee l fo r reduce d maintenance .
However, econom y migh t b e achieve d i f th e structur e was shielde d against extreme
temperatures wit h a temperature control envelope o r insulation layer .
2 Arrangemen t and appearance
Figure 1 illustrate s th e genera l arrangemen t fo r th e jointless stee l viaduct . Thi s

arrangement differ s significantl y fro m a viaduct wit h joints because : -
1) Span s ma y be longe r du e t o continuity ove r supports .

2) Pier s ma y b e mor e slende r becaus e brakin g an d tractio n force s ar e
transmitted t o abutments.
The jointless viaduc t i s show n wit h tw o continuou s longitudina l stee l plat e

girders, wit h composite cros s girders. Th e depth ha s been set at about 3.5m in order
to provide a 2m heigh t o f soun d barrie r togethe r with a reasonable dept h o f ballast .
Jointless steel viaducts for railway 16 9
The span s o f u p to 50m have been base d o n thi s depth o f plate girder.
The design principles demonstrate d here are applicable t o viaducts with bo x
girders.
3 Girder desig n
3.1 Temperatur e cycle s

BS5400 and TRRL 696 provide temperature data primarily for highway bridge s with
girders located below slabs. Th e jointless viaduct in Figure 1 has markedly different
thermal properties and considerable economy migh t be achieved by research into the
thermal behaviour of specifi c structura l form s with differen t coating s and envelopes ,
and b y calculation s base d o n firs t principle s a s i n Hambl y (1978) . I n orde r t o b e
conservative, i n the absence of specific information , this example is based on data in
BS5400.
Three distinc t type s o f temperatur e cycl e affec t th e fatigu e o f welde d joints
during th e 12 0 year design life:-
1) On e extrem e cycl e durin g th e 120-yea r desig n lif e betwee n th e extrem e

maximum and extreme minimum temperatures .
2) 12 0 annual cycles betwee n summe r and winter temperatures .
3) 36 5 x 12 0 daily cycle s betwee n da y and night temperatures.
Extreme temperature s are given in BS540 0 fo r 12 0 year return period. Annua l and
daily cycle s ca n b e estimate d fro m TRR L 696 , fro m whic h Figur e 2 i s copie d
showing histogram s o f dail y temperatur e changes fo r a stee l bo x girder . Th e
combined histogram of Figure 2 approximates to a Rayleigh distribution with a r.m.s.
value at about 1 3 degrees. Hambl y et al (1982 ) sho w tha t a Rayleigh distribution of
cyclic loadin g cause s the same fatigue damage as uniform cycles o f loading o f about
80% o f th e significant valu e (averag e of highes t 1/3 ) which i s equivalen t t o about
110% o f th e r.m.s.: i.e . 1 4 degree s here . Mor e detailed calculation s can also b e
done fo r annual and daily ranges , but are unlikely t o affect th e conclusion.
The overall forc e on th e viaduct cross-section i s caused by th e change in the
average effectiv e bridg e temperature . Howeve r th e stres s rang e at an y joint als o
includes th e locked-i n componen t fro m temperatur e differences. Usin g BS5400 ,
TRRL 76 5 and TRRL 696 , the viaduct in Figur e 1 might be expected to experience
the following cycles: -
Cycles Temperature changes

effective differenc e d T
deg C de g C de gC
Extreme 1 70 + 2 = 0 9 0
Annual 120 50 + 2 = 0 7 0
Daily 44,000 14 + 1 = 4 2 8
PLAN CROSS-SECTIO N
Figure 1 Genera l arrangemen t fo r a jointles s stee l viaduc t
Figure 2 Dail y r a n g e s o f e f f e c t i v e bridg e t e m p e r a t u r e o f s t e e l bo x bridge s (fro m T R R L 6 9 6 )

Jointless steel viaducts for railway 17 1
3.2 Therma l fatigue

Assuming coefficien t o f therma l expansion a= 0.0000T 2 pe r degree Centigrade an d
Young's modulus E=200,000N/mm , th e temperature ranges of dT=28 deg C induce
2
stress range s ds of
=
ds dTaE
= 2 8 x 0.00001 2 x 200,00 0 = 6 7 N/mm 2
Assuming Clas s F welds, BS5400:Ptl 0 Figure 14 indicates that for a stress range of
67 N/mm th e permissibl e numbe r o f cycle s i s abou t N= 2,000,000 . Her e th e
2
number o f dail y cycle s i s n= 44,000 : henc e th e fatigu e damag e from dail y cycle s
would be: -
D = n/ N = 44,000/2,000,00 0 = 0.0 2
Which i s negligibl e a s compare d t o th e permissibl e damag e o f D = 1.0. Simila r

calculations for the annual and extreme cycles of temperature also indicate negligible
fatigue.
3.3 Liv e load fatigu e

A preliminar y estimat e o f liv e loa d stresse s fo r acceptabl e fatigue desig n ca n b e
obtained from BS5400:Ptl0 paragraph 9.2. Assumin g R U loading the live load stress
range coul d b e u p to
ds = kx k k k k ds
2 3 4 5 G
as given i n BS5400:PtlO . Wit h Class F wel d details the equation becomes:-
ds = 1. 0 x 1. 0 x 2.1 9 x 0.89 x 1.2 7 x 4 0 = 10 0 N/mm 2
3.4 Ultimat e thermal force

It is assumed here that the viaduct has spans of 50m. R U loading o n both tracks will
cause a sagging momen t in each girder at midspan of about 20MNm, when th e span
is loaded, an d hogging momen t of -lOMNm, whe n adjacen t spans are loaded. Thu s
the rang e o f momen t wil l b e 30MNm . Wit h stres s rang e limite d b y fatigu e t o
ds = 100 N/mm , a s above, the section modulus Z of each girder will need to be about
2
Z = M/d s = 30/10 0 = 0.30m 3
If th e sectio n i s assume d t o be 3.5 m deep wit h radiu s of gyratio n r = 1.4 m it will

have momen t of inerti a I and area A of
I = 0.3 0 x 3.5/ 2 = 0.52m 4
A = 0.52/1.4 = 2
0.27m 2
If th e viaduct i s fixe d t o th e abutment s whe n at the middl e o f it s rang e of effectiv e

bridge temperature , the maximu m ris e of 3 5 de g C will caus e a compression stres s
f an d forc e P o f
f = 3 5 x 0.00001 2 x 200,00 0 = 8 4 N/mm 2
P = 8 4 x 0.2 7 = 23M N
Thus the abutment must resist a longitudinal forc e from two girders of 46MN (4,60 0
tons), pushin g o r pulling.
The maximum compression stres s of 8 4 N/mm wil l contribute to buckling o f
2
the viaduct span s a s a whole and to lateral buckling o f compressio n flanges .
4 Abutmen t desig n
It is show n above that the abutments must resist longitudinal force s of 46MN in both
directions. Th e abutmen t i n Figur e 1 is show n as a n open bo x o f diaphrag m walls
of abou t 25m x 30 m with depth 8m . I f the walls are 2m thick they will have weight
50MN and will enclose soi l of weight 80MN. I f the soil has angle of frictio n 3 5 deg
the maximu m slidin g resistance would be
Gliding = (5 0 + 80 ) tan(35°) = 10 0 MN
If th e soi l o f densit y 0.02MN/m ha s coefficient o f passive resistanc e of K p = 6.0 ,

3
then th e a wall o f widt h 30 m and depth 8 m could mobilis e a passive forc e of
F
P assive = 6
-° -
x 0 2 0x
(8 2/2
> x 3 0 = 11 5 MN
The safet y facto r against slidin g would be
SF = (10 0 + 115 )/46 = 4. 7
The maximum working coefficien t o f earth pressure would then be about K = Kp/S F
= 6/4. 7 = 1. 3 s o that the maximu m movement , accordin g to Hambly and Burland
(1979), woul d be o f th e order of 0.00 1 o f wal l depth , i.e . abou t 8m m (onc e in 12 0
years). Mor e precise calculation s coul d b e undertake n usin g finit e elements .
5 Pie r design
It i s anticipate d tha t th e complete d viaduc t wil l b e fixe d t o th e pier s i n orde r t o

dampen noise . I t is possibl e tha t the viaduct woul d be supporte d o n sliding/rollin g
bearings durin g constructio n an d the n fixe d t o th e pier s afte r th e viaduc t ha s bee n
attached t o th e abutments.
Under the action of longitudinal force s fro m braking and traction the jointless
Join tless steel via du cts for railway 17 3
viaduct will provide a stif f lin e of reactio n to the abutments, whil e the piers wil l b e
relatively flexible . Consequentl y ther e is no point in makin g th e piers stiff , an d their
flexibility wil l b e advantageou s i n accommodatin g th e effect s o f an y variation s i n
temperature alon g a structure . Th e tota l desig n longitudina l load s ar e smal l a s
compared t o th e temperatur e load o f 46M N calculate d above .
6 Reference s
Burke, M . P . (1990 ) "Integrate d bridg e desig n i s o n th e rise. " Moder n Stee l

Construction AISC , Jul y 1990.
Hambly, E . C . (1978) . "Temperatur e distribution s an d stresse s i n concret e
bridges"; Th e Structura l Engineer, 56A , Ma y pp 143-148 .
Hambly, E . C . an d Burland , J . B . (1979) . "Bridg e Foundation s an d
Substructures"; He r Majesty's Stationer y Office .
Hambly, E . C , Edwards , A . J . , Kohli, C . & Miller , B . L . (1981) . "Fatigu e
considerations fo r ocea n towage" . OT C 4163 , 13t h Offshor e Technolog y
Conference, Houston .
Hambly, E . C . & Nicholson , B . A . (1990) . "Prestresse d bea m integra l bridges" ,
The Structura l Engineer, Vo l 68 , No 23 , December .
Hambly, E . C . (1991) . "Bridg e Dec k Behaviour" , 2n d Edition , E & F N Spon ,
London.
Hambly, E . C . (1992) . "Integra l bridge abutmen t detail s i n practic e and theory",
Seminar o n Bridg e Desig n fo r Durability , Transpor t Research Laboratory ,
Crowthorne.
TRRL 69 6 (1976) , "Bridg e temperatures estimated fro m th e shad e temperature" , by
M. Emerson , Transpor t and Road Research Laboratory, Crowthorne .
TRRL 76 5 (1977) , "Temperatur e difference s i n bridges : basi s o f desig n
requirements", b y M Emerson , Transpor t and Roa d Researc h Laboratory,
Crowthorne.
R E P O R T O N SESSIO N FOU R
S. SHANMUGA M
The firs t thre e paper s o f thi s sessio n discusse d th e Danish , Frenc h an d Italia n
experiences wit h continuou s an d integral bridges.
Ejgil Veje briefly describe d some o f the recent major Danish bridg e projects,
which ar e mainly o f long span continuou s construction . Fo r a suspension bridg e h e
said tha t the y hav e designe d a 270 0 m lon g continuou s girde r betwee n th e tw o
anchor blocks, th e longest jointless lengt h eve r achieved i n bridge design. H e adde d
that suc h large spans have necessitated th e developmen t o f bearings an d expansio n
joints t o accommodat e the increase d movements involved .
Jean Mulle r showe d interestin g slide s describin g th e prestresse d concret e
continuous bridge s buil t i n Franc e an d how the y cope d wit h therma l effects .
Professor Cabrera described various types of continuous viaducts built in Italy
in bot h precas t and cas t in place , explainin g th e desig n an d constructio n processe s
involved. I n response t o a query about the longitudinal bo x girder being constructed
in two halves , th e Author said that each half was cas t separately and the units wer e
connected i n th e transvers e an d longitudina l direction s afte r positioning , an d tha t
there wer e n o problems .
In introducin g hi s paper , M r Paterso n referre d t o severa l continuou s arc h
bridges buil t betwee n 185 0 and 1900 ; continuous spa n bridge s constructe d in th e
1920s an d prestresse d concret e beam bridge s buil t i n th e 1970s . H e reporte d that
a smal l numbe r o f integra l bridges wer e buil t i n th e 1960 s and 1970s , al l o f whic h
have performe d very well.
He briefl y discusse d th e content s o f th e propose d Departmen t o f Transpor t
Advice Not e o n Desig n fo r Durabilit y whic h i s intende d t o giv e guidanc e o n goo d
practice. I t will be accompanie d by a Departmental Standard which will wherever
possible requir e al l bridge s t o b e continuou s ove r intermediat e pier s an d full y
integrated with abutment s for bridge lengths u p t o 7 0 m. I t will also includ e burie d
structures, an d arc h bridge s bot h i n concret e an d masonry , wher e the y stan d t o
benefit fro m reduced maintenance and enhanced durability . H e conclude d tha t the
re-establishment o f arc h bridge s ha s no w brough t th e whol e developmen t t o ful l
circle.
Dr Owens introduced the paper on Jointless Steel Viaducts and demonstrated
that fo r length s u p t o 20 0 m stee l viaduct s fo r hig h spee d railway s ca n perfor m
satisfactorily without expansio n joints a t either abutment. Fo r such stee l viaducts to
be compatibl e with welded railway track during temperature change , he pointed ou t
that massiv e anchore d abutment s would be required . H e suggeste d tha t econom y
might b e achieve d i f th e structur e was shielde d agains t extrem e temperatures.
PART FIV E
LONG-LENGTH
CONTINUITY
15 T H E D E S I G N AN D P E R F O R M A N C E
OF CONTINUOUSL Y REINFORCE D
CONCRETE PAVEMENT S
F. FUCH S
Belgian Road Research Centre, Brussels, Belgium
R. DUMON T
Ministry of Infrastructure an d Transport , Wallonia,
Belgium
Abstract
The pape r describe s th e us e an d behaviou r o f continuou s
reinforced concret e pavement s i n Belgium, i n particular th e
variants tha t hav e bee n use d i n constructin g CR C pavement s
at overpasse s withou t interruption .
Keywords : Bridge , Dec k surfacing . Continuous , Reinforce d
concrete.
1 Continuousl y reinforce d concret e pavement s i n Belgiu m
The continuousl y reinforce d concret e (CRC ) techniqu e ha s

been use d o n a larg e scal e i n Belgium , particularl y fo r
motorway constructio n sinc e th e earl y seventie s (ref . 1 ).
Over 1 8 million m o f CR C pavement s wer e constructe d sinc e
2
1970.
The mai n advantag e o f thi s typ e o f pavemen t i s th e
absence o f transvers e joints , whic h make s i t possibl e t o
construct lastin g pavement s requirin g virtuall y n o
maintenance. Crackin g i s controlle d b y continuou s
reinforcement locate d midwa y betwee n th e to p an d botto m o f
the sla b an d designe d i n suc h a wa y tha t a fin e networ k o f
cracks (averag e crac k spacin g betwee n 1 and 3 m) i s create d
which does not imperil the good performance of the pavement.
The mai n factor s influencin g th e crac k patter n ar e :
the percentage of longitudinal reinforcement related to

the yiel d poin t o f th e stee l an d t o th e tensil e
strength o f th e concrete ,
- bon d strengt h betwee n th e concret e an d th e steel .
In Belgiu m w e us e deforme d bar s wit h a yiel d poin t o f

500 MPa . Th e percentag e o f longitudina l stee l i s 0.6 7 %
with respec t t o th e cros s sectio n o f th e 2 0 c m thic k
concrete pavemen t (1 6 m m diamete r bar s wit h a spacin g o f
15 c m ) . Initially , fro m 197 0 t o 1976 , th e percentag e o f
longitudinal stee l wa s 0.8 5 % (1 8 m m diamete r bar s ever y
1 5 c m ) , but after a survey of the first motorways constructe d
it was reduced t o 0.67 % mainly i n order to increase averag e

178 Fuchs and Dumont
crack spacin g an d obtai n a more homogeneou s crac k pattern .

The reinforcement consist s eithe r o f all-welded fabri c
fully assembled i n the factory, or is composed of transvers e
bars provide d wit h support s welde d t o the m i n th e factory ,
the longitudina l bar s bein g subsequentl y fixe d t o thes e
transverse bar s o n th e constructio n sit e b y clippin g th e
ends o f th e supports . Th e transvers e bar s ar e placed a t a n
angle o f 60 ° wit h th e longitudina l bar s i n orde r t o
counteract the tendency fo r cracks to develop a t transvers e
bars perpendicularl y t o th e centr e lin e o f th e road .
The performance of most CRC pavements constructed sinc e
1970 can be rate d a s excellent, especially whe n considerin g
the littl e maintenanc e required . A surve y recentl y
conducted b y th e Ministr y o f Publi c Work s an d th e Researc h
Centre o f th e Cemen t Industr y ha s show n tha t ther e i s
practically n o corrosion o f the longitudina l reinforcement ,
the los s i n cross sectio n o f th e stee l bein g generally les s
than 1 % afte r te n t o twent y year s o f servic e (ref . 2) <
It ha s als o turne d ou t tha t thi s typ e o f pavemen t i s
able t o resis t subsidenc e i n th e embankment s o f bridg e
approaches, the structure and evenness of the pavement being
restored b y injectin g a cemen t grout .
The onl y defect s tha t hav e occurre d systematicall y
are :
- constructio n join t failur e durin g ho t summers , wit h

some case s o f blow-u p o f th e pavemen t du e t o th e poo r
quality o f th e concret e i n the vicinit y o f th e joint ,
- punch-out s a t th e oute r edg e o f th e pavemen t o n som e
recently constructe d motorways , du e t o inadequat e
drainage a t the interfac e betwee n th e pavement an d th e
road bas e an d th e poo r qualit y o f th e sub-bas e
material.
2 Continuousl y reinforce d concret e road s a t overpasse s
Interrupting CR C pavement s a t overpasse s raise s problem s ;
- o f constructio n : becaus e o f th e clos e spacin g o f

engineering structures on Belgian motorways, concreting
has to be stoppe d a t frequen t interval s an d transitio n
zones have to be made in bituminous materials, which i s
not a ver y rationa l workin g method ;
- o f cost : constructing ancho r abutment s an d transitio n
zones a t eac h engineerin g structur e i s expensive ;
- o f managemen t : wherea s CR C pavement s nee d littl e
maintenance fo r mor e tha n twent y years , joint s an d
bituminous surfacings requir e intervention s ever y fiv e
to seve n years .
There ar e three variants i n constructing CR C pavement s

at overpasses .
Continuously reinforced concrete pavements 17 9
30.00 10.00 10.00 10.00
i*
u IT ir
u u u L.
Expansion join t
Fig. 1 Interruptio n o f a CR C a t a n overpas s approach .
a) CR C o n overpas s approache s
At structure s no t designe d t o resis t force s applie d b y CRC ,
the pavemen t i s interrupte d b y a n ancho r abutment . Thi s
abutment i s locate d a t a distanc e suc h tha t th e foundatio n
of th e structur e i s shielde d fro m thrust s produce d b y
expansions o f th e concret e pavement .
The transition zon e between the anchor abutment and the

structure i s constructe d i n jointe d reinforce d concret e
slabs.
This procedure makes it possible to reduce stresses and
strains i n th e bridg e joint , bu t retain s al l th e
disadvantages involve d i n interrupte d concreting .
b) CR C o n conventiona l structure s
Frame bridge s
The stiffnes s o f thi s typ e o f structur e enable s th e CR C
pavement t o be continue d o n th e bridge a s fa r a s the desig n
engineer ha s allowe d fo r a permanen t loa d b y tha t pavemen t
(500 k g / m fo r 2 0 c m o f CRC ) an d a horizonta l forc e
2
equalling hal f tha t loa d (25 0 k g / m fo r 2 0 c m o f C R C ).

2
The bridg e ha s t o b e constructe d wit h a n extr a ± 1 0 %

of concret e t o resis t thos e forces , bu t thi s i s partl y
compensated b y th e possibilit y t o omi t th e floatin g slabs .
Girder bridge s
With spa n length s greate r tha n ± 2 0 m fram e bridge s ar e n o
longer cost-effective, and CRC pavements ar e constructed o n
girder bridges .
A significan t horizonta l forc e i s applie d b y th e
pavement to this type of structure, which i s less rigid tha n
a fram e bridge . Thi s horizonta l forc e i s reduce d b y
incorporating betwee n th e CR C an d th e waterproofin g syste m
of th e structur e :
- eithe r a doubl e polyethylen e shee t reducin g th e

horizontal forc e t o ± 2 0 % o f th e vertica l forc e
applied b y th e CRC ,
- o r sliding sheets reducing the horizontal forc e to ± 10
% o f th e vertica l forc e applie d b y th e CRC .
180 Fuchs and Dumont
6.00
i - CR C pavemen t (2 0 cm )
Expansion join t \
Lean concret e x
\ Stabilize d
-Sub-base x
embankmen
x t
\
10.00 X
Detail
Protective stri p ( 1 cm )
Expanded polystyren e
Fig. 2 CR C o n fram e bridges.
Road
Canal o f th e Intercepting
Espierre" ' ditc h
Detail CRC _ 0.2 0

Sliding shee t
Waterproofing laye r _ 0.01 5
Concrete r . 45 0 Bridg e dec k 0.11
Precast sla b _ 0.0 7
Precast girde r 195/8 4
Fig. 3 CR C o n girde r bridges.

Continuously reinforced concrete pavements 181
c) C R C on/in bowstrin g bridge s
The loa d superimpose d b y a concret e pavemen t o n th e bridg e
can b e reduce d i f the CR C pavement i s made a n integra l par t
of th e bridg e structure .
Neoprene bearing s betwee n th e transvers e girder s an d
the precas t slab s underlyin g th e CR C pavemen t allo w easie r
movement o f th e bridg e wit h respec t t o tha t pavement . Th e
pavement sla b i s lai d wit h a slip-for m paver .
Centre spa n
Fig. 4 CR C i n bowstring bridges .
3 Reference s
Y Dechamps , R D e Paepe , P Dutro n - Belgia n experienc e wit h

continuously reinforced concrete pavements. International
Conference o n Concret e Pavemen t Design , Purdu e Univer -
sity, 319-334 , 1977.
K Verhoeve n - Cracking an d corrosio n i n continuousl y con -
crete pavements. 5th International Conference on Concrete
Pavement Desig n an d Rehabilitation , Purdu e University ,
201-209, 1993 .
16 E X P E R I E N C E I N T H E U S E
OF CONTINUOU S WELDE D
RAILS O N TH E
NETHERLANDS RAILWA Y
NETWORK
G. BIE R
Ingenieursbureau, Nederlandse Spoorwegen ,
Utrecht, The Netherland s
Abstract
Modern track s consis t o f continuous welde d rails . Tempera -
ture change s introduc e force s i n the rail. At the ends of
the CW R there ar e anchor lengths , in which thes e force s
are reduce d t o nil by a number o f sleepers. Becaus e o f the
elasto-plastic resistanc e behaviou r o f the sleepers i n the
ballast, bi g displacements o f the rail end s occur , for
which specia l expansio n device s ar e needed.
Bridges unde r th e track giv e extr a stresse s i n the
rail, cause d b y both temperatur e elongatio n an d shortenin g
of th e bridge deck , an d by traction o r braking force s fro m
trains runnin g ove r th e bridge. Thes e effect s as k for
reduced bridg e dec k lengths , and minimum stiffnes s of
bridge bearing s an d foundation .
Keywords; continuou s welde d rail , ancho r length , expansio n
joints, bridg e dec k lengths , bridge bearing' s stiffness .
1 Wha t i s CWR on embankmen t
One o f the big changes i n Permanent Wa y engineering ha s

been th e introduction o n a larg e scal e o f continuou s
welded rai l (CWR) . Toda y mos t o f Netherlands Railway' s
network i s fitted wit h thi s lo w maintenance typ e o f track .
When yo u get rid of the fishplated rai l joint s tha t
were ther e ever y 1 8 , 24 or 30 meters, the question i s how
the mechanic s o f the system work .
For an y piece o f rail somewher e i n the middle o f the
CW-track w e can see that i t is in equilibrium betwee n the
two surroundin g ones . I t follows tha t i n this middl e
section o f the rail stresse s ar e being introduced . No
extra longitudina l forc e i s needed fro m belo w i.e . from
sleeper an d ballastbed .
This situatio n change s whe n we approach th e end of the
CWR section . Ther e mus t b e an end at CWR at places wher e
the interna l rai l forces , caused b y the CW, cannot be
withstood. Thi s ca n be the case a t complex swit h & cros -
sing spots , or before narro w curves , mine settlemen t area s
and, particularl y i n the Netherlands, at movable bridges .

by Taylor & Francis, 2 Park Square, Milton Park, Abingdon, Oxon, OX1 4 4RN. ISBN: 0 419 19030 9.
184 Bier
N = N - Tx
x
=
u
N= EAo c A t 2 l EA
a
x
2 l CA
T = 0
a
continuous welded rails x
AO-2.4 N/mm 2
/°C. For At = 40°C-—N - 700 kN per rail ; A<7* 100 N/mm 2
Fig.l. Distributio n o f temperatur e forc e i n CW R
At th e en d o f th e C W rai l w e kno w fo r sur e tha t ther e i s

no longitudina l force . A piec e o f rai l a t littl e distanc e
from th e rai l en d wil l b e pushe d awa y b y a bigger forc e
coming fro m th e CW-sid e tha n fro m th e sid e o f th e
expansion joint . Thes e piece s i n the s o calle d anchor - (o r
breathing-) lengt h ca n onl y b e brough t int o equilibriu m b y
an extr a forc e delivere d b y th e sleeper . S o here , th e
sleeper doe s hav e t o resis t certai n longitudina l force s
from th e CWR .
Since th e longitudina l force s i n the CW R ca n reac h
values o f 50 0 t o 70 0 kN s pe r rai l (stresse s u p t o 100 -
N / m m ) , i t i s impossibl e t o sto p thes e force s b y on e
2
sleeper o r fastening . A numbe r o f sleeper s togethe r hav e

to resis t th e CW R forces . B y doin g so , every sleepe r i s
pushed throug h th e ballas t befor e i t come s t o it s maximu m
resistance value .
This mean s tha t a t th e en d o f th e CW R th e rai l ca n
undergo quit e som e variatio n i n length , which reduce s th e
internal rai l forc e t o zer o a t th e end . Bu t thi s i s accom -
panied b y suc h displacement s a t th e end , tha t specia l rai l
expansion device s (adjustmen t switches ) ar e t o b e used .
The movemen t i n th e ballas t occur s whe n rai l force s
reach a certai n level . The bigge r th e temperatur e change s
are, th e mor e resistanc e i s needed. Frequen t movement s o f
the sleeper s throug h th e ballas t ar e unfavourabl e fo r th e
packing o f th e ballas t an d thu s influenc e bot h ballas t
resistance (longitudina l an d lateral ) an d th e track' s
level an d alignment . Whic h reduce s ridin g qualit y an d
increases th e nee d fo r trac k maintenanc e work .
One othe r mechanic s aspec t i s th e ris k o f trac k
buckling a t highe r temperatures .
Considering a straigh t ba r tha t w e pu t unde r longitudi -
nal pressure , we kno w tha t a t a certai n leve l o f interna l
stress, th e ba r buckles . Holding thi s ba r i n th e middle ,
raises th e critica l longitudina l bucklin g forc e t o th e
Continuous welded rails in the Netherlands 18 5
square. Holdin g th e bar on two more points , again incre -

ases th e critical bucklin g forc e resistance . The miracle
of CW R is that norma l sleepere d trac k can , under specifi c
conditions, resis t th e forces tha t occu r i n our climatica l
circumstances.
Rail temperature s can , in the Netherlands, as probabl y
is the case i n the United Kingdom , var y betwee n minu s 2 5
and plu s 5 5 degrees Celcius . We always faste n th e rails a t
the sleeper s a t the same, so-calle d neutra l (stres s free )

temperature. Sinc e trac k bucklin g i s more critica l the n
high tensio n stresse s i n the rail are , this temperatur e i s
chosen abov e th e mean valu e o f the temperature limits . At
NS i t is 25°C.
2 Wha t rol e doe s the bridge play ?
When w e loo k bac k o n the situation i n the middle o f the

CWR, we see the influence o f the bridge comin g t o the
track i n the form o f an extra longitudina l displacemen t of
the bas e o f the ballastbed. Thi s lead s to forces comin g
from ballastbe d t o sleeper an d finaly rail , resultin g i n a
disturbance o f the continuous CW R stress level . At the
movable en d of the bridge, rail stresse s increase , wherea s
over th e fixed bearin g are a the y ge t reduced.
Three limit s ar e to be considered (i n order o f impor -
tance) :
1. th e track bucklin g resistance .

2. th e rail (tension ) strength .
3. displacement s i n the permanent way construction ,
especially o f importanc e whe n the y occu r often .
Since th e cross sectio n o f the bridge dec k i s so much
bigger tha n tha t o f the rail, n o specific attentio n i s
paid t o stresses i n the deck. The more however , we mus t
consider th e forces in , and displacements o f the bridge
deck supports , i n particular whe n rubbe r bearin g block s
are used .
And, no t to forget, do the same fo r the rail supports .
3 Th e load s
For th e interaction betwee n trac k an d bridge mainl y two

characteristic longitudina l loads , must b e considered :
1. Temperatur e effect s
Caused b y the climatical influence s th e bridge deck s
expand o r shorten. One can distinguish betwee n the
variation ove r th e year (slow ) and over th e day
186 Bier
(quicker, bu t stil l o f stati c nature) . Thes e effect s

occur mainl y i n th e unloade d trac k situation .
P= 19.7 7 k N I
7T
N = EAO d A t
-200
37 m
At sp s = 40° C
At de k = 10° C
Fig.2. Temperatur e effect s
2. Tractio n (maximu m 33, 3 kN/m' ) an d breakin g force s (max -

imum 2 0 k N / m ' ) .
Hi TT) HTT TT) HTYT TYT
Fig.3. Tractio n an d brakin g effect s

These force s ar e quasi-static t o even dynami c o f charac-

ter. Especiall y lon g freigh t train s ca n give big forces.
German survey s hav e show n tha t th e biggest forc e occur s a t
the las t 0, 2 second s perio d o f the braking manoeuvr e o f a
train. Ever y ca r will sto p a t the same tim e whic h leads ,
at lon g trains , to a very bi g peak loa d o f dynamic kind .
In flexibl e bridge s thi s ca n cause ver y bi g forces i n the
continuing rai l an d displacements o f bridge t o rail and
rail t o embankment .
4 Betwee n trac k an d bridge
Between rai l an d bridge ther e i s always a more o r les s

elastic element . At ballasted tracks , favoured b y perma-
nent way engineers, it is mainly th e ballastbed tha t act s
like a n elasto-plastic medium . Th e resistance value s o f a
ballastbed ar e depending o n the shape o f the stones, the
ballast dept h belo w the sleeper, the type o f sleeper, the
fastening properties , and, very important , the compactio n
of th e ballast. The compaction i s much lowe r afte r mainte -
nance operation s o r when th e ballast move s frequentl y a
lot.
source: UIC leaflet

720 R
= consolidated ballast = 100%
after tamping
\^\ after tamping + shoulder compaction
| 11 1 after tamping + overall compaction
23 4 5 (mm )
movement
F i g . 4 . los s o f latera l ballas t resistanc e du e to main-

tenance operation s
When an y type o f direct rai l fastenin g i s to be used ,

investigation mus t b e done t o the elasto-plastic behaviou r
of it . The advantage o f direct rai l fastenin g ca n be the
more constan t forc e an d displacement properties . Les s
188 Bier
favourable i s th e greate r stiffnes s an d reduce d displace -

ment capacity . Thi s result s i n general i n les s possibili -
ties fo r makin g CW R ove r bridges .
5 Rai l expansio n joint s
In orde r t o compensat e bigge r displacemen t betwee n tw o

following rai l sections , specia l device s hav e bee n develo -
ped. The y ar e characterise d b y a join t tha t crosse s th e
rail a s longitudina l a s possible . Thus , tw o tongue s ar e
shaped.
Like i n turnouts , the passin g whee l firs t run s a t on e
tongue, an d the n graduall y start s t o ru n a t th e othe r one .
This takin g ove r goe s mor e smoothly , th e flatte r th e angl e
of th e joint s t o th e trac k directio n is .
Although thes e construction s ar e muc h mor e expensive ,
they ofte n canno t compet e wit h th e cost s o f bridg e dec k
joints.
Another point , however is , that thes e device s nee d muc h
more inspectio n an d maintenance . Als o the y have , i n gene-
ral, shorte r lif e times , which mean s tha t mor e ofte n tha n
the surroundin g rails , they hav e t o b e replaced .
Expansion joint s interrup t th e CWR , an d this , a s w e
have mentione d before , creates ancho r lengt h areas , with
all thei r disadvantage s suc h a s movement, wear, disturban -
ce o f trac k leve l an d alignement , an d s o on .
1950
R A I L SLIDIN G A R E A
CROSS SECTIO N A - A
190 Bier
6 Proli s compute r program m
Since, i n th e Netherland s w e hav e onl y littl e difference s

in level , many bridge s ar e designe d wit h direc t railfaste -
ning becaus e o f it s lo w constructio n depth . Becaus e thos e
are mor e influencin g th e interactio n betwee n bridg e an d
track, a lo t o f calculation s hav e bee n done . The mathema -
tical possibilitie s hav e grow n much , an d s o we hav e final -
ly see n th e developmen t o f a ne w computerprogramm , calle d
Prolis. I t was develope d b y tw o student s o f th e Technica l
University o f Delft , accompanied b y severa l peopl e fro m
Netherlands Railways . Base d o n th e finit e element s princi -
ple man y propertie s i n the desig n ca n b e varie d t o calcu -
late severa l differen t variants .
7 UI C 7J1 8
By UIC , the Internationa l Unio n o f Railwa y Administrati -

ons, a workin g grou p (calle d 7J18 ) was established , wit h
the ai m t o brin g fort h a desig n leafle t fo r railwa y brid -
ges. Bot h peopl e fro m th e Bridg e Department s (BR , DB, SNCF)
and Permanen t Wa y Department s (N S an d SNCB ) too k par t i n
this group . After severa l year s o f exchang e o f thought s
last Januar i a genera l agree d proposa l wa s offere d t o th e
charging Bridg e commite e (UI C 7 J ) . Discussion i n th e
Permanent Wa y committe e (UI C 7 G ) , is stil l unde r way ,
where especiall y th e matte r o f maximu m rai l force s an d
displacements i s stil l i n view. Estimate d limit s fo r
additional railstresse s cause d b y bridge s ar e 7 2 N/mm fo r
2
pressure an d 9 2 N/mm fo r tension . Movemen t o f bridg e

2
decks unde r CW R shoul d sta y unde r 5 mm .
8 Experience s a t th e N S networ k
From th e earlie r day s o f th e railwa y w e ha d man y bi g rive r

crossing bridges , all fitte d wit h ballastles s track , an d
with us e o f expansio n joints .
After th e secon d worl d war , an d especiall y fro m th e mi d
sixties, man y ne w viaduct s an d bridge s hav e bee n built . I n
our mor e an d mor e dens e populate d country , th e design s
became mor e an d mor e complicated . No t onl y straight , bu t
also curve d wit h pil e pattern s sometime s ver y complicated .
Trough bridge s skewein g a t very shar p crossin g angle s t o
the infrastructur e beneath , enabl e fre e crossin g o f traf -
fic flows , but as k sever e desig n work .
Until th e mi d eightie s man y bridge s wit h direc t rail -
fastening hav e bee n designe d an d built . With those , afte r
ten t o twent y fiv e year s o f experience , we hav e gaine d
knowledge abou t wha t i s possible .
Not alway s th e experience s wer e fre e choice . Fo r exam -
ple, jus t i n th e ol d cit y o f Delft , we hav e a n 800 m lon g
5=2 S= 3 G== > £=3 5= 2 £= 3

Support arrangement 1 "
1 C — =3 1
22. 0.9 0.3 2 \
2,
\ / \ / \ / \ \ /
\
\
\
\
CM 86.3 | \
4
7
86.6
86.6
86.3
85.4
61.4 /
60.8 /
61.6 /
(Si
60.8 /
I 85.
to
Fig.5. Longitudina l rai l force s a t lon g bridge s
viaduct, doubl e tracked . Abou t hal f th e number o f train s

is passin g a t a spee d o f 10 0 to llOkms/hour . The othe r
trains pas s a t lo w speed an d are braking fo r a stop i n
Delft Centra l Station . At the end, closest t o this stati -
on, ther e i s a curve o f abou t 550 m radius . The viaduc t
exists o f 1 7 sections o f 55 m length each , and the rail s
are directl y fastene d t o the decks. When th e viaduct was
put i n use, there wa s a rail expansio n join t a t ever y
bridge joint . Thos e however , alread y afte r shor t tim e gav e
big trouble s wit h maintenance , especiall y i n the narro w
curve. A solutio n wa s found , withou t compute r facilities ,
but stil l a n analytical mode l showe d tha t i t was possibl e
to mak e CW R over th e viaduct. I t was necessary t o creat e
anchor lengt h ove r bot h en d abuttments, in order t o pre -
vent th e firs t dec k bein g pushe d fro m it s piles.
All calculation s tha t hav e bee n don e fo r the so calle d
Delft's Regula r Railfastenin g sho w i n fact tha t a dec k
length o f 5 5 m i s to long . However , base d o n many year s o f
experience, we have practicall y considere d th e 55 m to be
possible. Probabl y ther e i s a favourabl e effec t o f som e
relaxation i n the railfastening .
9 Today' s practic e
Our mor e recen t practic e howeve r i s that we try to get

ballasted track s designe d o n the bridges enablein g th e
bridge designe r t o cross ove r dilatatin g length s o f 9 0 m
for concret e an d 60 m fo r steel bridges . These dec k leng -
ths ar e advantageous, i n particular i n the complicate d
situations earlie r mentioned .
For th e PW engineer especiall y th e smoothening o f the
transition t o the adjacant embankment , and the possibilit y
of continuin g maintenanc e operation s ove r th e bridge ar e
of grea t advantage .
When however , bridg e deck s hav e t o be so lon g tha t
expansion joint s ar e needed anyhow , the fixed positio n o f
those joint s i n direct railfastenin g i s preferable.
17 C O N T I N U O U S A R C H E S
D.R. THOMPSO N andG.W . CLAR K
British Rail Research, Derby, UK
Abstract
Masonry arc h bridge s ar e constructe d entirel y withou t th e
use o f movemen t joints . Thi s pape r look s a t th e natur e
of masonr y arc h bridge s an d highlight s thei r mai n
attributes. Th e mai n strength s an d weaknes s o f thi s typ e
of structur e ar e considere d i n th e contex t o f th e
continuous bridg e deck .
Keywords; Masonry , Arch , Viaduct .
1 Introductio n
Masonry arche s ar e on e o f th e oldes t form s o f bridg e

construction. Th e arc h wa s use d b y man y o f th e ancien t
civilisations bu t th e Roman s ar e notabl e a s many fin e
examples o f thei r arche s hav e survive d an d som e stil l
carry traffi c 200 0 year s afte r the y wer e buil t [1].
Masonry arche s o f al l type s wer e extensivel y use d i n
the constructio n o f railways , particularl y i n Britain ,
and man y o f thes e structure s ar e stil l i n use today .
Masonry arche s ar e use d i n structure s rangin g fro m simpl e
culverts t o mil e lon g tunnel s an d viaducts .
Masonry arche s ma y b e constructe d fro m man y differen t
materials suc h a s brick , ston e o r mas s concrete .
An arc h i s a pur e compressio n structure ; i t carrie s
loads b y mean s o f individua l block s (voussoirs ) pressin g
against eac h other . Thi s for m o f constructio n i s
fundamentally stable , an d ha s prove d ver y durable .
This pape r wil l loo k briefl y a t th e wa y i n which th e
arch works . I t will sho w ho w individua l arche s ma y b e
strung togethe r t o for m lon g bridge s withou t th e nee d fo r
construction joints . Finally , a compariso n o f th e arc h
and bea m fo r continuou s bridge s wil l b e made .
2 Ho w a n arc h work s
Figure 1 show s a n idealize d arc h wit h th e mai n element s

194 Thompson and Clark
Spandrel Wall Parape t Wall
Figure 1 : Idealise d arc h
detailed. Arche s ma y be constructed fro m simpl e singl e

elements suc h a s shaped stone s o r bricks (masonry ) o r
from mas s concrete . Th e vast majorit y o f the arches o n
British Rai l (BR ) are constructed fro m masonry . Masonr y
(and mas s concrete ) i s characterised b y it s inability t o
sustain tensil e forces .
The arche s considere d her e are , by definition, pur e
compression structure s wit h n o ability t o resist interna l
tensile forces . Th e use of reinforced concret e o r stee l
in a n arch woul d lea d t o a structur e wit h ver y differen t
characteristics an d is outside th e scope o f this paper .
The mai n loa d bearin g elemen t o f an arch i s the
barrel. Th e barrel ca n be considered a s a serie s o f
elements (Figur e 2 ) . Eac h elemen t support s it s own dea d
weight an d the dead weigh t o f the fill abov e it . I t is
held u p by component s o f the forces exerte d o n it by
the element s o n eithe r sid e o f it . Consequentl y th e dea d
load stresse s fro m th e self weigh t o f the fill an d the
barrel materia l ar e carried b y the barrel. Th e centre o f
action o f the stresses i n the barrel i s called th e thrus t
line.
These dea d loa d stresse s ar e crucial t o the stabilit y
of th e arch. The y ac t in a simila r manne r t o the pre -
stress i n a pre-stressed concret e bea m i n that the y
enable a material wit h n o tensile capabilit y t o carr y
loads whic h woul d otherwis e produc e tensil e stresses . The
Continuous arches 19 5
Figure 2 : Force s o n a barre l elemen t
greater thi s 'dea d loa d prestress ' the greate r th e liv e

load capacity , unles s th e limitin g compressiv e stres s o f
the masonr y i s exceeded . Th e positio n o f th e thrus t lin e
need no t b e coinciden t wit h th e centroi d o f th e barrel ,
although, sinc e th e thrus t lin e mus t remai n withi n th e
barrel thicknes s a t al l times , the close r i t i s t o th e
centroid, th e mor e loa d th e arc h ca n carr y fo r a give n
barrel cros s sectio n an d material . Th e geometr y o f th e
arch i s importan t here , a s som e shape s (e.g . elliptical )
lead t o th e developmen t o f a thrus t lin e tha t i s clos e t o
the surfac e o f th e barre l eve n unde r dea d loa d only .
If a n externa l forc e i s applie d t o a n arch , the n th e
thrust lin e alter s it s positio n i n th e barrel , t o
maintain th e stati c equilibriu m o f th e structure . A s th e
thrust lin e move s awa y fro m th e neutra l axi s o f th e
barrel, bendin g moment s ar e introduced . Th e furthe r th e
thrust lin e move s the n th e bigge r th e bendin g moment s
will become . A s th e thrus t lin e move s outsid e th e middl e
third o f th e barre l thicknes s tensil e stresse s tr y t o
develop a t th e opposit e surface . Thi s lead s t o tensil e
cracking o f th e masonr y an d a redistributio n o f th e
compressive stres s acros s th e remainin g par t o f th e
section. A s th e cracke d zon e grows , the are a o f materia l
available t o resis t th e thrus t i n th e barre l decrease s
and th e compressiv e stres s increases . Thi s increas e i n
local stres s ma y caus e crushing . Onc e al l o f th e materia l
across the section has either cracked or crushed no

further increase in the loading at that section can be
sustained and a plastic hinge forms. Due to its shape,
an arch can still function with three such hinges in most
cases. When a fourth hinge forms, however, the arch
becomes a mechanism and collapse occurs.
The precise points at which the hinges develop depends
on the geometry of the arch and the position of the
loading. It is the development of these hinges which
controls the ultimate capacity of the arch in most cases.
The behaviour of a real arch is complicated by the
interaction between the barrel and the surrounding fill.
As the arch barrel deforms, the whole structure (both
barrel and fill) distorts as a homogeneous whole.
Movements in the fill cause stresses within the fill
itself and between the fill and the external surface
(extrados) of the arch barrel. This interaction between
the soil and the structure makes the analysis of arches a
complex problem. The behaviour of arch fill material is
not fully understood. In practice it is often difficult
to determine the nature of the fill and its physical
properties.
The durability of the arch comes partly from its
reliance solely on compression for its strength. This
means that the arch is inherently stable, even if
individual elements deteriorate or crack. The compressive
stresses within the arch barrel will tend to close up
many cracks. Other faults can be accommodated by means
of a re-alignment of the thrust line and changes in the
shape of the arch.
3 Ho w a viaduct work s
Figure 3 shows the form of a typical viaduct. A masonry

arch viaduct is a series of simple arches on piers. This
form of construction allows a bridge of unlimited length.
Each span can be different from those abutting it both in
span and shape.
The way in which the viaduct behaves under load
depends on the slenderness of the piers. Stocky piers
can carry both vertical and horizontal loads directly to
the foundations. This means that deflections from one
span are localized and are not carried over to the
adjacent spans. In many cases the piers are quite
slender and cannot carry the entire horizontal load from
one arch span. In such cases the load is also carried by
adjacent spans as the loaded arch barrel deflects and
changes shape. This pier sway away from the load causes
upward deflections in the adjacent spans, which are
resisted primarily by the dead weight on those spans but,
also by the pier stiffness. Consequently the live load
Arch Change s Shap e t o Maintai n Stabilit y
Any Numbe r o f Span s
One o r Mor e Pier s Ma y Transmi t Longitudina l Force s
Figure 3 : A typical viaduct
deflection of one arch span causes reduced movements in

the adjacent arch spans. This effect is repeated in the
next arch and so on until the movement, and therefore the
stresses, decay away completely.
The inability of slender piers to carry all of the
out-of-balance horizontal forces means that such viaducts
are difficult to construct one span at a time and are
susceptible to progressive collapse if a span is removed.
For this reason stocky piers are usually built into such
viaducts every five spans or so.
Masonry (and mass concrete) viaducts have no need of
movement joins. Both loading and thermal movements, are
absorbed by the structure in the form of changes in the
shape of the individual arch barrels.
The analysis of masonry arch viaducts is a very
complex operation. Problems caused by the fill/barrel
interaction and the behaviour of the fill, for a single
arch, are compounded by the ability of the supports of
each individual arch to move independently. The failure
mode of masonry viaducts is considerabley more complex
than single arch spans; the collapse mechanism frequently
involving hinges in several spans and piers. The effect
of the spandrels is also more pronounced in multispan
arches than in single span arches [ 2 ] .
British Rail Research have developed the Masonry Arch
Finite Element Analysis (MAFEA) [ 3 ] program to cope with
the comple x problem s presente d b y arc h analysis . MAFE A

has bee n i n use for single span s fo r some time . Thi s
program will , i n time, be extended t o cover multi-spa n
structures.
As wit h singl e spa n arches , viaducts ac t as a
homogeneous entity . Consequentl y longitudina l forces ,
such a s braking an d traction, ar e distributed ove r th e
whole lengt h o f the structure an d dissipated. I n the
case o f viaducts wit h stock y pier s thes e force s ca n be
expected t o pass directl y int o th e supports.
Figs 4 and 5 show th e Culloden an d the Harringwort h
viaducts. Th e Culloden viaduc t i s good exampl e o f the
continuous bric k masonr y viaduct s tha t ca n be seen all
over th e UK. Not e th e enlarged spa n t o carry th e
structure ove r a road. Th e Harringworth viaduc t show s
the considerabl e lengt h o f structur e tha t ca n be
achieved.
4 Compariso n wit h continuou s bea m structure s
Bridges base d o n the simply supporte d bea m principl e ar e

probably th e most commo n for m o f modern construction .
This typ e o f structur e i s very simpl e t o design an d to
analyse. Joint s ar e used t o divide th e bridge int o
simple element s fo r analysis. Thi s compel s deflection s
to occu r a t predictable place s throughou t th e bridge.
Recent experience , particularl y wit h highwa y bridges ,
has show n tha t thes e joint s ca n be difficult t o maintai n
and inspect . Th e decay o f the material aroun d joint s
often lead s t o expensiv e repair s an d even th e need t o
replace th e structure completely .
Arches d o not need movemen t joints . Consequentl y the y
avoid thi s weakness , which i s inherent i n the design o f
some moder n bridges .
In moder n bridg e decks , expansio n joint s ar e ofte n
used t o remov e th e stresses tha t woul d b e induce d i f
temperature-related expansion s an d contractions wer e
restrained. However , longitudina l brakin g an d tractio n
forces, mus t the n b e carried acros s th e these joints .
Anchoring o f the bridge dec k agains t longitudina l
forces i s normally achieve d b y means o f a smal l numbe r o f
fixed bearings . Thi s mean s tha t th e braking an d tractio n
loads ar e concentrated a t a few points, leadin g t o hig h
local stresses . I n a masonry (o r mass concrete ) viaduc t
the resistanc e t o longitudina l force s i s distribute d
throughout th e lengt h o f the structure an d there i s no
concentration o f forc e a t joints .
Fig 5 : Harringwort h
5 Discussio n
Large masonr y arc h bridge s ar e no longe r bein g

constructed fo r a numbe r o f reasons. The y ten d t o be ver y
heavy an d consequentl y the y ma y require larg e
foundations. Arche s ma y be expensiv e t o construct an d
require labou r wit h skill s tha t ar e not readily availabl e
in today' s market . Masonr y arche s ar e not amenable t o
rapid constructio n an d must b e built in-sit u du e to the
need fo r centering. Th e former disadvantag e coul d b e
largely overcom e b y use of mass concret e instea d o f
masonry bu t the latte r applie s equall y t o both materials .
The curve d arc h profil e reduce s headroo m adjacen t t o the
springing. Arche s ar e not suitable fo r larg e singl e span s
unless a larg e constructio n dept h i s available an d
consequently lon g arc h bridge s hav e t o be constructed a s
multi-span viaducts . Thu s th e designer mus t accep t a
large numbe r o f intermediat e supports .
The difficult y i n analyzing arc h structure s an d in
predicting thei r performanc e ca n also dete r som e
engineers. Designer s hav e becom e use d t o the more
rigorous solution s availabl e fo r more moder n bridg e
forms.
However ther e ar e also considerabl e advantage s wit h
the arc h for m o f constructio n whic h shoul d no t be
overlooked. The y hav e prove d thei r durabilit y o n
railways ove r th e las t 15 0 years an d have bee n i n use for
canals an d roads fo r considerably longer . The y hav e
proved t o be lo w maintenance structures , despit e th e fac t
that man y ar e now carrying considerabl y highe r load s the n
they wer e originall y designe d for . They hav e a good loa d
carrying capabilit y an d they ca n be built t o for m
viaducts o f any lengt h withou t th e need fo r movemen t
joints.
Compared t o modern constructio n forms , masonry arche s
are simpl e i n concept. Thi s combine d wit h thei r lo w
maintenance need s an d proven 12 0 year plu s lif e
expectancy shoul d lea d t o favourabl e whol e lif e costing .
The advantage s liste d abov e ar e achieved b y virtue o f
the combinatio n o f the simplicity o f for m an d materials.
Masonry arc h structure s d o not require reinforcin g stee l
or comple x component s suc h a s bearings. Thi s simplicit y
of for m an d materials mean s tha t the y ar e not likel y t o
be pron e t o unexpected failur e mechanism s a s is the cas e
with reinforce d an d prestressed concrete . I n addition ,
maintenance i s simple an d no specia l components ,
materials o r techniques ar e needed .
6 Conclusion s
The followin g point s ma y be drawn i n favour o f masonr y

arch bridge s fro m th e abov e analysis :
(a) The y d o no t requir e movemen t joints .

(b) The y hav e prove n structura l reliabilit y ove r mor e
than 15 0 years .
(c) The y ca n b e constructe d fro m material s tha t ar e
inherently durable .
(d) Th e liv e loa d stresse s ar e distribute d aroun d th e
structure, an d no t concentrate d a t a point, o r a
small numbe r o f points .
(e) The y ar e inherentl y stable .
And agains t masonr y arc h bridges :
(f) The y requir e comple x analysis .

(g) The y ar e ver y heavy .
(h) The y ca n onl y b e use d fo r relativel y shor t span s
or fo r a serie s o f shor t spans .
The arc h i s on e o f th e oldes t bridg e form s an d i s

often ignore d b y today' s bridg e designer s o n th e ground s
that th e initia l constructio n time , an d therefor e cost ,
rule i n favou r o f bea m bridg e solutions . I t i s th e
authors' opinio n tha t th e inheren t durabilit y o f th e arc h
form woul d ofte n lea d t o lowe r whole-lif e cost s fo r a n
arch structure . Arche s ar e aestheticall y mor e pleasin g
than beam s an d ca n b e mad e no t onl y t o blen d i n with th e
landscape bu t sometime s t o enhanc e it . At a tim e whe n
engineers ar e bein g criticise d fo r payin g insufficien t
attention t o th e aestheti c aspect s o f thei r design s w e
should loo k agai n a t th e arc h a s a n inherentl y durabl e
and elegan t solutio n t o man y desig n problems .
7 Reference s
Melbourne, C an d Wagstaf f M (1993 ) Loa d test s t o

collapse o f thre e large-scal e multi-spa n brickwor k
arches. Pro c o f Bridg e Managemen t 2 , Thomas Telford ,
London.
Smith N.A.F . (1993 ) Th e Roma n Bridge-builder : Som e
aspects o f hi s work . Th e Structura l Engineer , Vo l 71 ,
No 9 , 4 May 1993 .
Woolfenden P.A . (1993 ) Modellin g th e masonry arch :
improving moder n bridg e assessmen t usin g nonlinea r
finite-element softwar e packag e (MAFEA) . Pro c o f
Bridge Managemen t 2 , Thomas Telford , London .
REPORT O N SESSIO N FIV E
B.P. P R I T C H A RD
Messrs Fuch s an d Dumon t wer e unfortunatel y unabl e t o atten d and , a s a result ,

there wa s n o presentatio n o r discussion o f thei r interesting paper .
Mr Bie r described hi s considerabl e experienc e o f continuou s welde d rails o n
Dutch railways . Dutc h polic y afte r the war was t o get rid of rail joints an d now 90%
of track is CWR. H e mentione d th e very large forces involved (700 kN per rail) and
the en d ancho r requirements, whic h us e 6 0 m o f trac k an d underg o u p t o 6 0 m m
movement.
Track bucklin g i s mor e critica l tha n hig h tensio n an d clos e spacin g o f th e
stabilising sleepers i s therefore necessary. Bridge s carrying track cause problems an d
some movemen t betwee n trac k an d bridge i s accommodate d by suitabl e ballasting .
Rail expansio n joint s ar e often required . Batterin g of rai l joints i s ofte n a proble m
and thi s ca n be reduced , a s ca n be th e noise , b y glued-i n rails.
Mr Bie r wa s questione d abou t th e larg e brakin g effect s o n trac k an d
confirmed tha t recent research has shown much larger loading tha n the Dutch cod e
had stipulated . Ther e wa s som e furthe r discussio n abou t th e U K us e o f shoc k
transmission unit s t o alleviat e thi s brakin g b y sharin g i t ou t amon g joint-separate d
elements.
Mr Thompso n provide d ampl e evidenc e o f th e benefit s o f joint-free bridg e
construction by describing Britis h Rail's long standing reliance on multi-span masonry
railway arches . H e describe d thei r abilit y t o absor b larg e expansion/contractio n
movements b y arc h deformation, als o their ready ability t o absorb large traction and
braking forces . Undoubtedl y ther e i s a cas e fo r thei r continuin g us e b y th e bridg e
engineer, particularl y with BR' s experienc e o f 12 0 year s plu s lif e wit h minimu m
maintenance problems .
There was som e discussio n about the use o f simple mass concrete arches with
plastic sheetin g separator s t o provid e th e joint s normall y forme d betwee n th e
masonry units . Th e pioneerin g wor k o f Bolto n Universit y wa s mentioned .
PART SI X
RETROFITTING
18 S T U D Y O N JOINTLES S
BRIDGE CONSTRUCTIO N
IN J A P A N
H. TAKAN O
Engineering Department, Metropolitan Expressway
Public Corporation, Tokyo, Japan
Abstract
The pape r reviews Japanes e experience with embedded an d combined-typ e
jointless method s o f bridge construction . Thei r main characteristic s are describe d
and illustrated.
Keywords: Japanes e bridges; embedde d joints ; combined-typ e jointles s methods ;
joint types .
1 Introductio n
In Japan , wit h the rapi d construction of nationwid e roa d networks afte r th e

Second Worl d War, the numbe r o f bridges increase d a t a n accelerate d pace.
Among these , thos e buil t i n th e earlie r periods wer e predominantl y o f th e simpl e
beam type , usuall y use d i n combinatio n wit h expansion devices . However , i t
became apparen t that the expansio n device s wer e susceptibl e t o damag e du e t o
increasing traffi c volume a s well a s increasing number s o f heav y vehicles, resultin g
in deterioratio n of roa d conditions an d becoming a sourc e o f vibration an d noise
pollution. A t th e sam e time , frequen t repai r of expansio n device s congeste d
traffic. Accordingly , governmen t authoritie s an d organization s responsibl e fo r
construction an d maintenanc e o f publi c roads , including th e Ministr y of
Construction, Japa n Highwa y Publi c Corporation , Metropolitan Expressway Public
Corporation an d Hanshin Expresswa y Public Corporatio n began makin g
concerted effort s t o develo p a constructio n method whic h eliminate s expansio n
joints b y pavin g o r partially combining bridg e sections , thereb y achievin g a
continuous an d smoot h roa d surface. Thi s resulte d i n th e developmen t o f various
construction method s fo r existin g shor t spa n bridges, generall y referre d to a s
jointless constructio n methods .

by Taylor & Francis, 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN. ISBN: 0 419 19030 9.
208 Takano
Jointless method s ar e roughly classified a s follows:
— Expansio n absorbing type

Embedded type joints Expansion dispersing type
Jointless 1
— Expansio n inducing type
methods —
— Mai n beam connection method
Combined type jointless method s — Floo r slab connection method
Crossbeam connection method
To develo p thes e technologie s int o systemati c technical standards, the Committee

for Stud y an d Research on Jointles s Construction Methods fo r Existing Bridges
(Chairman: Minor u Fujiwara, Director , Bridge Structure Department , Public
Works Researc h Institute, Ministry of Construction ) was organize d an d a research
project launched .
2 Embedded-typ e joints
There ar e many types o f embedde d joints which hav e been use d fo r actua l bridge
construction sinc e 1982 . Followin g experimenta l installation i n various projects ,
these joints were introduce d throughout the country . Th e existin g technolog y o f
embedded typ e joint, however , ha s not provide d a perfect constructio n method
because i t i s incapabl e o f avoidin g th e occurrenc e of crack s o r ruts. Nevertheless ,
the technologica l reliabilit y ha s significantl y improve d sinc e i t sufficientl y reduce s
damage t o th e joints an d the adjoinin g area .
This method ha s been use d i n a number of bridge construction s and now i s
considered t o hav e become practical . Th e embedded-typ e joint s nee d no t b e
permanent structur e an d can be installe d whe n roads are repaved. Thus , the
embedded-type joint s nee d no t be permanen t structure an d can be installe d whe n
roads ar e repaved. Thus , the embedded-typ e joint s ar e sufficiently usefu l i f the y
have durabilit y similar to tha t of roa d pavement. Mino r damage to th e joints will
not b e considere d seriou s i f it cause s no problem s t o vehicular movement. Th e
embedded-type joints , particularly , offer suc h advantage s as smooth drivin g an d
easy maintenance when the surfac e of embedde d joint are a i s connecte d
completely leve l with the adjoinin g roa d surface.
Based o n suc h evaluation , man y types o f embedde d joints became
technically an d commerciall y feasible fo r use i n actua l projects under certain
conditions. Deman d increasingl y grew for government organization s responsibl e
for roa d administration to compil e embedde d joint technologie s int o standard
manuals. Instructions , tentatively named "Embedded-Typ e Joint Design and
Construction Manual (proposed)," have been prepared.
In developin g thes e manuals , a questionnair e surve y was conducte d among
related organization s to identify ho w embedded-typ e joint s were being used i n
actual roa d projects. Result s showe d tha t the numbe r of lanes employin g suc h
Jointless bridge construction in Japan 20 9
joints a s o f th e surve y had reached approximately 3,000. Th e embedded-typ e

joints ar e classified mainl y into thre e types accordin g to thei r function: expansio n
absorbing, expansio n dispersin g an d expansion inducin g types . Th e expansio n
absorbing typ e i s adapte d easily t o absor b expansion an d contractio n by using sof t
material fo r th e entir e pavement. Thi s typ e o f pavement, however , i s vulnerable
to rut s worn by heav y traffic .
Concerning th e expansio n dispersin g type , sof t pavemen t materia l is no t
used, bu t a shee t i s lai d unde r the pavemen t t o dispers e expansio n an d
contraction. Thi s typ e prevent s ruts worn by heavy traffic while the pavemen t
adjacent t o th e embedde d typ e joint i s liabl e t o crack . I n short, the expansio n
absorbing an d dispersin g type s ar e contradictory in structura l characteristics .
The expansio n inducin g typ e is a n embedded-type join t designe d t o induc e
expansion an d contraction to a cut-of f joint specificall y create d in th e pavement .
This structure , therefore , limits crackin g to suc h specific cut-of f joint are a i n th e
pavement.
The embedded-typ e join t i s a n expansion devic e t o absor b deformation
that occur s in th e joint structur e b y utilizing the pavement' s material propertie s
and th e embedde d simpl e join t mechanism . Tabl e 1 and Figure 1 indicate
features an d conceptua l diagram of majo r embedde d typ e joints. I t should b e
noted tha t embedded typ e joints, du e t o difference s i n physica l propertie s and
Table 1 Majo r feature s o f embedded-typ e joint s
Type Features
Expansion - Soft pavement materials are used, and the deformation of such
absorbing typ e materials absorbs the expansion and contraction or revolution
at the extreme end of main beam.
- Difference i n level is likely to occur between the embedded
joint area and other areas because these two pavements pos-
sess different physical properties.
Expansion - Basically, the pavement structure's deformation is utilized. A

dispersing type sheet which is called a Shea r Layer is placed between the
pavement structure and floor slab. Expansion and contraction
are dispersed to the entire pavement structure by the shear
layer's shearing deformation.
Expansion - Deformation is induced to a cut off-joint provide d in

inducing type the pavement structure.
- This method is suitable for small bridge construction. Whe n
deformation grows, however, the cut off-joint become s a weak
point and the cut off-joint are a is susceptible t o damage.
210 Takano
Pavement Length of embedded joint Pavement
Soft pavement structure / Floor slab Adhering surface
Expansion gap Backup material
< Expansion <t

contraction |
(Expansion-absorbing type )

Setdement Elastic Settlement
section section section
Shear layer Adhering surface

Floor slab

\ . con u
(Expansion-absorbing type )
Cut off-joint /
2f Adhering surface
Floor slab
<
Expansion Sc. I
(Expansion-induction type )
Figure 1 Conceptua l diagram of major embedded-typ e joints

Jointless bridge construction in Japan 21 1
pavement configuration , ar e affected b y various condition s i n varying degree , suc h

as traffi c conditions , typ e o f bridge structure , maintenanc e requirements , and
frequency o f traffi c congestion .
3 Combine d typ e jointless metho d
The combined-typ e jointles s metho d join s togethe r the member s o f adjacent

beams t o transmi t the temperature-induce d displacement towar d the bridg e axi s
to th e adjacen t beams. Th e beams connecte d by combined-typ e joint s behav e i n
the sam e way a s the continuou s beam s d o when a n earthquak e occurs. Th e
beam's behaviou r i n respons e t o live load show s various patterns ranging from
behaviour clos e t o th e continuou s bea m t o on e clos e t o th e simpl e beam . Th e
beams connecte d by embedded-typ e joint s behav e i n th e sam e manne r as th e
original simpl e beam s afte r completio n o f jointless processes , an d the numbe r of
spans ar e limited whe n the combined-typ e jointles s metho d i s applied . A
combined-type jointles s metho d shoul d b e selecte d accordin g to bridge types .
Table 2 indicates th e combined-typ e jointles s method s an d applicabl e bridg e type .
Application o f th e combined-typ e jointles s metho d accompanie s structura l
modification, whic h result s i n a larger constructio n scale a s compare d with the
embedded-type join t method . Th e numbe r o f construction s employing combine d
type jointless method s i s no t large .
Figure 2 show s a basic classificatio n o f combined-typ e jointles s methods .
Table 2 Combined-typ e jointles s method s an d bridge type s
Methods Bridge type
Main beam connection method

Steel beam
Floor slab connection method
PC beam
Crossbeam connection method
Acknowledgement
The autho r deeply appreciate s Mr M Ito , Professor o f Saitam a University

and Professor Emeritus of Universit y o f Tokyo , fo r giving him a n opportunit y t o
present thi s pape r of th e colloquium .
212 Takano
Floor slab connection area Pavement
*3 Connecting reinforcing bar
- Flange connection plate
Embedded joint
a) F l o o r s l a b c o n n e c t i o n m e t h o d ( s t e e l b e a m )
Pavement Floor slab
/ l
End crossbeam ^
> PC steel bar

Filling concrete -
Floor slab connection area b ) C r o s s b e am c o n n e c t i o n m e t h o d (PC b e a m )

Pavement
/
Floor slab
] /
/
Connecting
plate
Stiffener
/
I1 11
| Elastic bearing
c) M a i n b e a m c o n n e c t i o n m e t h o d (stee l beam )
Figure 2Basi c classification of combined-type jointless methods

19 S E M I - I N T E G R A L BRIDGES* :
A CONCEP T WHOS E TIM E
HAS COME ?
M.P. BURKE , Jr
Burgess & Niple Ltd, Columbus, Ohio, USA
Abstract
The Stat e o f Ohi o ha s conceive d an d i s developin g a semi-integra l bridg e
concept tha t ha s enable d i t t o exten d th e applicatio n rang e of jointless bridg e
decks, eve n t o bridge s no t normall y considere d fo r thi s typ e o f construction.
Although thi s concep t ha s a wide applicatio n range , it possesses a number o f
unusual characteristic s that must b e recognize d an d be provide d fo r i f semi -
integral bridge s ar e to satisf y al l functiona l requirements . A discussio n o f
these characteristic s is th e primar y focus o f thi s paper.
1 Introductio n
The Stat e o f Ohi o ha s bee n buildin g continuou s concret e slab bridge s wit h
flexible integra l abutment s fo r 6 decades . I t has bee n buildin g continuou s
steel bea m an d girde r bridges wit h flexibl e integra l abutment s fo r 3 decades .
Although thi s typ e o f constructio n is no w routin e i n Ohi o fo r mos t bridges ,
there wer e a numbe r o f exception s fo r suc h applications . Fo r example, bridge s
skewed greate r than 30 degrees , continuou s bridge s longe r tha n 300 fee t
(91.4 m) , bridges wit h wall-typ e abutment s an d bridges wit h stub-typ e
abutments o n rigi d foundation s wer e routinel y provide d wit h movabl e dec k
joints a t abutments . Mor e recently , however , Ohi o ha s conceive d an d i s
developing a semi-integra l abutmen t concep t tha t has enable d i t t o exten d th e
application rang e o f bridge s wit h jointles s deck s t o mos t moderat e size bridges ,
even t o thos e application s wit h exceptiona l characteristic s (Figure 1) . Thi s
semi-integral abutmen t concep t i s als o no w bein g used t o retrofi t existin g end -
jointed continuou s bridges .
* Although variou s type s an d size s o f integra l o r jointless bridge s hav e bee n

constructed fo r centuries , th e term s "integral bridge" and "semi-integra l bridge "
used i n thi s pape r refer to th e typica l smal l singl e an d multipl e spa n
continuous bridge s bein g used t o bridg e smal l stream s and roadwa y crossings.
They generall y constitut e th e bul k o f structure s being constructe d today.

214 Burke
Fig. 1 . A semi-integra l abutmen t concep t fo r

single an d multipl e spa n continuou s bridges .
Although th e semi-integra l desig n concept describe d i n thi s pape r has

expanded th e applicatio n rang e for bridge s withou t dec k joints, i t possesse s a
number o f unusua l characteristic s that must b e recognize d an d be provide d
for. Otherwise , applicatio n o f thi s typ e o f desig n may result i n bridge s tha t d o
not satisf y al l o f thei r functional requirements . A discussio n o f thes e
characteristics i s th e primar y focus o f thi s paper.
2 Basi c feature s
The basi c feature s o f th e semi-integra l abutmen t concep t illustrate d i n

Semi-in tegral bridges 21 5
Figure 1 include th e absenc e o f a bridge dec k joint; a superstructur e tha t

moves longitudinall y o n elastomeri c bearings almos t independen t o f rigi d
abutment foundations ; abutmen t members , including pilin g that can be
designed t o operat e well within th e usua l allowabl e stres s limits; a reduced
area o f structur e expose d t o passiv e pressur e and pressure which i s les s
eccentric wit h respec t to th e neutra l axis o f th e superstructure ; an d abutmen t
and end-diaphrag m configuration s whic h ar e simple t o design , simpl e t o
reinforce, an d relativel y simpl e t o construct.
Notice, however , tha t this desig n doe s no t eliminat e th e nee d fo r movabl e
joints. I n fact, i t double s thei r number since i n additio n t o a movable joint a t
the leve l o f th e bridg e seat , anothe r movable joint i s neede d betwee n th e
approach sla b an d approac h pavement. Ye t while doublin g thei r number, it
has reduce d thei r vulnerability t o substantia l maintenance . I f these joints fai l
to functio n a s desired , thei r failure will not hav e th e damagin g consequence s
that ha s com e t o b e associate d wit h malfunctionin g dec k joints an d joint
sealing systems .
The basi c features o f thi s desig n concep t hav e engendere d a number of
other structura l problem s whic h mus t b e considere d an d resolve d fo r eac h
semi-integral bridg e application . Thes e includ e th e contro l of behavior a t the
structure-backfill interface , design o f buried sealin g system s fo r movabl e joints ,
and placement o f concret e between surface s where sligh t differentia l
movements ar e probable.
3 Superstructur e restrain t
Of al l th e characteristic s o f th e semi-integra l bridg e concep t describe d i n thi s

paper, th e longitudinal , lateral , and vertical restraint of th e superstructur e ar e
the mos t important . Thi s typ e o f structur e shoul d no t b e considere d fo r desig n
unless it s designe r i s familia r with thes e characteristic s an d make s appropriate
design provision s t o accoun t for them .
3.1 Longitudina l restraint

As ca n be surmised b y viewing the semi-integra l abutmen t desig n show n i n
Figure 1 , the superstructur e whic h i s supporte d o n movabl e elastomeri c
bearings move s almos t independen t o f th e abutments . Tha t is th e reaso n why
this desig n concep t i s adaptabl e to bridge s wit h variou s type s o f rigi d
abutments. Fo r bridges withou t fixe d piers , i t receive s it s longitudina l restrain t
almost exclusivel y fro m source s no t normall y utilized i n bridg e desig n fo r thi s
purpose. Longitudina l restrain t comes fro m approac h slab-subbase friction ,
shearing resistanc e o f elastomeri c bearings, an d th e compressiv e resistanc e of
structure backfill . However , durin g col d weather , afte r the superstructure
contracts awa y from th e abutment s (an d awa y from th e backfill) , onl y th e
shearing resistanc e o f hal f o f th e bearing s an d th e frictiona l resistanc e o f th e
approach slab s movin g relativ e t o th e subbas e wil l be immediatel y availabl e t o
216 Burke
restrain th e superstructur e against externall y applied longitudina l forces . Fo r

this reason , it would b e desirabl e i f th e granula r backfill a t abutment s coul d b e
placed an d consolidate d durin g col d weather , o r at night durin g ho t weather s o
that th e backfil l coul d initiall y contribut e more restraint to supplemen t tha t of
the approac h slabs an d bearings fo r resistin g longitudina l forces .
Providing turn-bac k wingwalls cantilevere d fro m th e superstructur e in lieu
of straigh t wingwalls woul d provid e additiona l longitudina l restrain t by
mobilizing th e resistanc e o f backfill-wingwal l friction , o r for wingwalls wit h
irregular surfaces , the shearin g resistanc e o f th e backfill . Fo r longer multipl e
span structures , attaching the superstructur e to a free standin g pie r would b e
another way o f providin g additiona l resistanc e to longitudina l forces .
Generally, longitudina l resistanc e provided b y approac h slabs an d bearings
should b e sufficien t t o satisf y specificatio n requirement s respecting th e
resistance t o longitudina l forces . Fo r moderate earthquake forces, th e
resistance provide d b y the consolidate d backfil l shoul d provid e th e additiona l
longitudinal restrain t needed fo r moderat e length bridges , eve n durin g col d
weather. Fo r longer bridges , anchorag e to piers ca n provide th e extr a
longitudinal restrain t needed, eve n fo r large longitudinal forces .
3.2 Latera l restraint

Figure 2 show s a n elevatio n vie w o f a typical semi-integra l abutment . Notic e
that th e superstructur e of suc h structure s is separate d from the abutmen t by a n
essentially horizonta l movabl e joint a t the bridg e sea t an d vertical movabl e
joints betwee n th e superstructur e and transvers e wingwalls. Th e horizonta l
bridge sea t joint i s show n i n Figur e 1 . Th e vertical wingwall joint s ar e similar
except tha t only filler s an d sealer s ar e provided fo r unskewed bridges .
Consequently, th e abutment s o f thi s semi-integra l bridg e concep t functio n
essentially a s longitudina l guide s fo r th e superstructure.
Vertically, th e superstructur e is supporte d b y elastomeri c bearings i n th e
bridge sea t joint. Laterally , the superstructur e is supporte d by th e interaction
of th e superstructure , approach slab, an d backfill; t o som e exten t by th e
compressive resistanc e o f th e fille r i n th e latera l joints; an d t o som e exten t by
the shearin g resistanc e of th e elastomeri c bearings i n th e bridg e sea t joint.
For application s wher e substantia l latera l resistance is necessar y (suc h a s
skewed structure s [see below] , o r structures exposed t o strea m flow pressur e or
earthquake forces ) guid e bearing s ar e necessary an d thei r use i n th e wingwal l
joints o r elsewhere betwee n beam s i s recommended .
For superelevate d bridge s wher e th e bridg e sea t joint (Figur e 2) i s slope d
parallel t o th e dec k surfac e an d where th e elastomeri c bearings ar e als o
sloped, latera l guide bearing s ar e necessary to provide th e suppor t needed t o
resist th e latera l component o f th e superstructur e reaction. Otherwise , suppor t
bearings mus t b e se t o n leve l bridg e sea t surfaces.
When considerin g th e applicatio n situation s tha t have t o b e contende d wit h
in th e desig n o f semi-integra l bridges , th e us e o f guid e bearing s shoul d b e
considered routin e fo r mos t i f no t al l semi-integra l bridg e designs .
Fig. 2 . Semi-integra l abutmen t with transvers e wingwalls .
3.3 Rotationa l restraint

Based o n th e analysi s give n below , superstructure s of som e skewe d semi -
integral bridge s will , unles s restraine d by guide bearings , ten d t o rotat e i n a
horizontal plane . Thi s tendenc y wil l b e greate r for bridges wit h greate r skews.
Horizontal rotatio n will initiat e soone r fo r longe r bridges .
As superstructure s of semi-integra l bridge s expan d i n respons e t o risin g
ambient temperatures , superstructure elongation (AL ) will be resisted b y
backfill bein g compresse d a t abutments (Figur e 3). Forc e i s require d t o
compress backfil l an d thi s sam e forc e will restrai n superstructure elongation b y
inducing compressiv e stresse s i n th e superstructure . Whe n considerin g th e
relative compressibilit y o f backfil l an d a reinforced concret e superstructure ,
even thoroughl y consolidate d granula r backfill, i t shoul d b e clea r that almos t
all o f th e expecte d superstructur e elongation wil l occu r as compressio n o f
backfill. Onl y a sligh t amoun t o f compressio n wil l occu r in th e superstructure
as evidenc e b y a sligh t reductio n i n th e amoun t o f superstructur e elongatio n
that woul d hav e bee n eviden t ha d th e elongatio n bee n unresisted . Thes e
compressive stresse s ar e shown summarize d i n Figur e 4 a s th e resultant
longitudinal superstructur e compressive forc e P sec a Th e centralize d
p
location o f thi s resultan t force i s base d o n th e assumptio n tha t structur e

backfill i s homogeneous , an d tha t i t would b e uniforml y compresse d
218 Burke
L A L
P A S S I V E P R E S S U R EP D
Fig. 3 . Passiv e pressur e developmen t fo r semi-integra l bridges .
throughout th e widt h o f th e superstructure . Th e component s o f thi s resultan t

force agains t th e backfil l ar e th e norma l forc e du e t o passiv e pressur e P an dp
the latera l forc e P sec esi n e , o r i n simple r terms , P tan a I f lateral guide
p p
bearings fo r th e superstructur e are not provide d an d th e forc e P tan e i s no t

p
adequately resiste d a t th e structura l backfill interfac e (b y frictio n o f backfil l o n

superstructure end-diaphragm s [P tan S] or by th e shearin g resistanc e o f
p
backfill [P tan 0]) , differential movemen t a t th e structure-backfil l interfac e will

p
commence. Whe n considerin g th e shearin g resistanc e o f backfil l (P tan 0 ) o r p
the frictiona l resistanc e o f backfil l o n smoot h concret e surface s (P tan <5),

p
usually th e latte r force wil l b e foun d t o b e th e smalle r o f th e tw o an d i t will

govern behavio r a t th e structure-backfil l interface .
Since th e externa l force s ac t "on " both end s o f th e superstructur e of a semi -
integral bridg e (Figur e 4), the eccentri c longitudinal forc e component s P will p
tend t o rotat e the superstructur e towards th e acut e corner s of th e structure , or

for th e ske w shown i n Figur e 4, i n a clockwis e direction . Th e latera l forc e
components o n th e othe r han d (P tan 0 o r P tan S) will ten d t o resis t thi s
p p
rotation.
Using th e shearin g resistanc e o f a n idealize d granula r backfill an d th e
frictional resistanc e o f backfil l o n th e backfill-structur e interfac e surfaces , i t
can b e show n tha t superstructure s o f semi-integra l bridge s skewe d greate r than
Fig. 4 . Th e elongatio n an d rotatio n o f semi-integra l bridg e superstructures.
about 1 5 degrees wil l b e unstabl e unles s the y ar e provided wit h guide bearing s
at bot h abutments .
With respec t t o Figur e 4, th e abov e statemen t ca n be justified b y a short
computation a s follows :
Where: L Bridge dec k lengt h

= e Bridge ske w angl e
Total passiv e pressur e
FS
= Factor o f safet y
0 = Angl e o f interna l frictio n o f backfil l
6 :
Angl e o f structure-backfil l interfac e frictio n
For th e superstructur e of a skewe d bridg e t o b e stable , th e forc e coupl e

tending t o resis t rotatio n (P tan <SLco s e ) mus t b e equa l t o o r greater than th e
p
force coupl e tendin g t o caus e rotatio n (P Lsin e) , or;

p
P L si n e < P tan £Lcos e (1

p p )
Providing a facto r of safet y agains t rotation ;
P Lsin e < Pptan <SLco s Q / F S

p (2)
220 Burke
Since th e weight o f attache d approach slabs an d slab-subbase frictio n will

tend t o resis t movement , a safet y facto r o f 1. 5 seems sufficien t fo r thi s
situation. Insertin g this facto r i n equatio n (2 ) and simplifying ;
sin e < tan <Sco s e/1. 5

tan e < tan <S/1. 5
g < arc tan (tan <S/1.5 ) (3 )
Assuming tha t the angl e o f frictio n a t the structure-backfil l interface (S ) is

22 degree s (se e pag e 7.2-6 3 NAVFAC [1982] fo r granular backfill o n a smoot h
concrete surface) , equatio n (3 ) suggests tha t the bridg e skew angl e g must b e
equal t o o r less tha n 1 5 degrees t o b e stable . Fo r greater skews , i t i s likely
that rotatio n will be initiate d unles s guid e bearing s are provided a t both
abutments t o resis t th e force s inducin g suc h movement .
Other observation s ca n be mad e with respec t to equation s (1 ) and (3).
Although equatio n (1 ) indicates tha t some leve l o f passive pressur e must b e
generated t o caus e rotation ; equation (3 ) indicates tha t the ske w angl e a t
which rotatio n will be initiate d i s independen t o f both passiv e pressur e and
bridge lengt h an d directl y related to structure-backfil l interface friction.
The magnitud e o f guid e bearin g reactions is anothe r indication o f th e
potential fo r superstructur e rotation . Sinc e mos t o f th e therma l movement o f
a superstructur e will be paralle l to th e longitudina l axi s of a bridge, guid e
bearings shoul d b e place d paralle l to thi s axis . The n based o n th e lateral
force component s show n i n Figur e 4, the guid e bearin g reaction, which woul d
be norma l to thi s axis , is give n b y P^tan gcos q, o r in a simple r form , P sin a
p
Based o n thi s relationship , the require d capacity of guide bearing s for a

structure skewe d 3 0 degrees i s equa l t o 0.5P , o r one-half o f th e tota l passiv e
p
pressure. Fo r a 45 degre e skew , th e require d capacity equals mor e than 70

percent o f th e tota l passive pressure . Consequentl y i t i s clearl y evident tha t
neither th e frictiona l resistance (P tan 6) nor the shearin g resistanc e (P tan 0)
p p
of th e backfil l ca n resist force s o f thi s magnitude . Consequently , guid e

bearings shoul d b e provide d fo r structure s with larg e skews if a stabl e
superstructure an d a fully functiona l bridg e is t o be provided .
3.4 Vertica l restrain t

Due t o thei r jointless construction , many types o f integra l bridges ar e buoyant
when the y become submerged . Thi s i s tru e fo r many I-beam bridges an d som e
spread-box beam bridges.
The weigh t o f diaphragm s at abutments provide som e resistanc e to uplift .
But generally , som e positiv e desig n provision s mus t be mad e t o ensur e that
integral bridge s hav e a reasonabl e factor o f safet y agains t flotation. I-bea m
webs ca n be pierce d near top flange s b y 3-inch diamete r holes space d
uniformly throughou t the bea m length ; th e spac e between sprea d boxes ca n
also b e vente d b y placing 3-inc h minimu m diamete r horizontal vent duct s near
the to p flang e o f al l beams . Thes e duct s should pass completel y throug h the
Semi-integral bridges 22 1
beams fro m on e we b t o th e othe r and the y shoul d b e place d i n concrete

diaphragms o r be completel y encase d i n concret e to preven t floodwaters fro m
entering bea m voids; counterweight s coul d b e use d bu t thei r weight mus t b e
taken int o accoun t during beam design ; uplif t restraint s could b e provide d a t
pier bearings ; or some pier s ca n be integrall y constructe d with th e
superstructure t o ad d sufficien t uplif t restrain t to counterac t buoyancy.
In lieu o f vent holes , adde d weight , uplif t restraints , or integral pier
construction, th e us e o f th e mos t buoyan t structure s should b e restricte d to
those bridg e site s wher e th e highes t floodwate r level s ar e well belo w th e
superstructure.
4 Desig n aspect s
4.1 Movabl e joint s

As ha s bee n mentioned , th e semi-integra l bridg e with it s attache d approach
slabs ha s eliminate d bridg e dec k end-joints . I n their place, i t ha s incorporated
two othe r joint types , a movable joint a t the leve l o f th e bridg e bearings, an d
one a t th e pavemen t en d o f th e approac h slab. Althoug h doublin g thei r
number, th e desig n ha s minimize d thei r significance. Les s tha n desirabl e
performance fo r eithe r o f thes e joints will not hav e th e significantl y advers e
consequences tha t has com e t o b e expecte d with th e failur e o f bridge dec k
joints. Th e bridg e sea t joint an d cycl e contro l joint ar e the tw o joints tha t
have bee n provide d t o accommodat e the movemen t o f semi-integra l bridges .
In addition , on e mus t mentio n tha t rigid approac h pavements als o mus t b e
provided wit h effectiv e pressur e relief joints t o guar d semi-integral bridge s
from uncontrollabl e approac h pavement growth .
4.11 Bridg e sea t joint s

In th e desig n show n i n Figur e 1 , the troublesom e bridg e dec k joint ha s bee n
eliminated. However , i n it s place , a movable joint ha s been introduce d a t the
level o f th e bridg e seat . Corrosio n resistant elastomeri c bearings ar e provided
so tha t the superstructur e can move longitudinall y almos t independen t o f th e
rigid abutments .
The movabl e bridg e sea t joint mus t b e provide d wit h a durabl e elastomeric
seal sinc e i t i s burie d i n th e backfil l an d consequentl y i s no t accessibl e fo r
repair o r replacement. Otherwis e it s mos t importan t characteristic i s it s abilit y
to preven t backfill fro m being force d int o th e joint b y compresse d backfill . I t
would b e desirabl e bu t no t absolutel y necessar y for th e sea l t o b e watertight.
It mus t als o permi t unrestrained differential movemen t betwee n th e abutmen t
and superstructure , even fo r bridges with larg e skews, an d it mus t retai n these
characteristics fo r man y years without th e nee d fo r repai r o r replacement.
Although th e bridg e sea t joint sea l i s a n important aspect o f th e semi -
integral bridg e design , Ohi o ha s yet t o adop t a desig n tha t appears to fulfill al l
of th e necessar y functional an d durabilit y characteristics . A numbe r o f tria l
222 Burke
designs hav e bee n develope d an d used. Initially , standar d compression seal s

were employed . The n i t becam e apparen t that a reinforced elastomeri c shee t
type sea l wa s mor e functionally suitabl e fo r both squar e and skewe d
applications. Th e shee t sea l no w bein g use d i s a 3/32 inch thic k nylo n
reinforced neoprene . I t is attache d to th e bridg e by various means includin g
elastomeric ancho r rods in forme d recesses , stee l clam p bars with expansio n
anchors, washer s an d masonr y nails, o r bonding adhesives . I t remains to b e
seen whic h on e o r more o f thes e attachmen t methods wil l be adopte d an d
perfected b y Ohi o fo r thi s critica l joint.
4.12 Cycl e control joints

Semi-integral bridge s wit h attache d approach slabs lengthe n an d shorte n i n
response t o temperatur e and moistur e changes . Consequently , fo r suc h
structures, th e boundar y between approac h slabs an d approac h pavement
should b e provide d wit h cycl e contro l joints t o facilitat e suc h movement .
Otherwise, longitudina l cyclin g o f both structur e and approac h slabs ca n
damage bot h flexibl e an d rigi d approac h pavements.
At present , standar d pavement expansio n joints wit h compressiv e filler s ar e
being provide d fo r shorte r semi-integral bridges . Longe r bridges ar e bein g
provided wit h pavemen t pressur e relief joints (wid e joint s fille d wit h asphal t
concrete).
For semi-integra l bridge s buil t adjacen t to rigi d pavement, i t i s imperativ e
that the y als o b e protecte d from pressures due t o th e restraine d growth of
approach pavement . Effectiv e pressur e relief joints shoul d b e provide d fo r al l
semi-integral bridges , eve n th e shortes t bridges. Consequently , fo r bridge s
adjacent t o rigi d approac h pavement, tw o type s o f pavement joints are
required; on e t o facilitat e th e cycli c movement o f th e bridg e an d th e othe r to
protect th e structur e and cycl e contro l joints fro m the effect s o f pavemen t
growth. Design s by fou r Transportatio n Department s ar e illustrated by Burke
(1987). A s note d i n tha t reference, all design s presentl y bein g use d hav e thei r
limitations.
For longe r integra l bridges, Ohi o i s usin g wide pressure relief joints t o
serve bot h purposes . Sinc e integra l bridges an d semi-integra l bridge s ar e such
new conceptions , muc h additiona l developmen t i s neede d i f th e approach
pavement joints adjacen t to suc h structure s are to provid e al l o f th e necessar y
attributes tha t these joint s mus t hav e t o satisf y structur e requirements withou t
continuous maintenance .
4.2 Backfil l
Backfill fo r semi-integra l bridge s shoul d no t b e considere d a nuisance tha t has
to b e contende d wit h a s i s th e cas e with th e full y integra l bridge o n flexibl e
abutments. Instead , backfill shoul d b e recognize d a s a n integral part an d a n
important par t o f th e semi-integra l bridg e concept . A s i n th e cas e o f a
retaining wall supporte d b y sprea d footings o n soil , th e wall whe n properl y
designed wil l interac t compositely wit h th e soi l an d be adequatel y supporte d
by th e soi l bot h verticall y an d laterally . I n a similar fashion, th e superstructure

and backfil l o f semi-integra l bridge s for m a partially composite interactiv e
structure. I n thi s context , th e backfil l perform s multiple functions . Whil e rigi d
abutments provid e bot h vertica l and latera l support for th e superstructure , the
backfill supplement s thi s suppor t b y providing vertica l support fo r approach
slabs an d bot h longitudina l an d latera l support for th e superstructure . An d th e
ultimate succes s o r failure o f th e semi-integra l bridg e concep t will depen d t o a
great exten t upo n method s an d procedure s that are developed b y th e bridg e
engineering professio n t o enhanc e th e interactio n between th e superstructure
and backfill .
Since backfil l i s suc h a n integral par t o f a semi-integra l bridg e concept ,
every effor t shoul d b e mad e t o ensur e tha t it i s properl y selected, constructed ,
and maintained .
Backfill shoul d b e selecte d an d designe d t o hav e characteristic s suitabl e fo r
superstructure-backfill interaction ; it shoul d b e o f a compositio n tha t protects
it fro m erosio n an d i t shoul d b e protecte d from abov e b y full-roadway widt h
approach slabs . Fo r bridges wit h confine d drainag e (raised curbs , barriers,
parapets, etc.) , approac h slabs mus t b e provide d wit h curb s with a heigh t o f at
least 6 inche s o r more t o confin e roadwa y drainage and conduc t i t alon g
bridge approache s and awa y fro m th e backfill . Approac h roadway curb inlets
should b e considere d an d b e provide d i f necessar y to ensur e effectiv e drainag e
control.
An effectiv e subdrainag e syste m shoul d als o b e provide d i n th e backfil l
above imperviou s embankment s t o ensur e tha t the retentio n o f subsurfac e
water i s minimized .
Provisions als o shoul d b e mad e t o intercep t subsurface approac h roadway
drainage an d discharg e i t awa y fro m th e abutmen t backfill. Granula r subbases
should b e provide d wit h efficien t latera l drains to discharg e subbase drainag e
laterally t o embankmen t sid e slopes . Roadwa y underdrains must b e
terminated beyon d th e bridg e approac h slabs an d be provide d wit h lateral
drains t o embankmen t sid e slopes . Otherwise , underdrai n accumulations
should b e conducte d i n close d conduit s longitudinall y throug h the backfil l an d
abutments.
Finally, bridg e maintenanc e engineer s shoul d becom e familia r with semi -
integral bridg e characteristic s s o tha t they ca n properly appreciate the
importance o f backfil l superstructur e interaction and provide th e corrective
maintenance tha t such structure s must hav e i f the y ar e to provid e th e servic e
life tha t their design anticipated.
5 Constructio n aspect s
Unlike thei r jointed bridg e counterparts , semi-integral bridge s hav e concrete

and backfil l placemen t peculiaritie s tha t must b e recognize d an d be provide d
for durin g design . Placemen t o f end-diaphragm s an d approac h slab concrete
224 Burke
must b e limite d t o th e exten t tha t the therma l movement o f th e superstructure

during an d shortl y afte r concret e placement i s minimize d t o preclud e
damaging freshl y settin g concrete . Thi s ca n be don e b y limiting concrete
placement t o thos e day s when rapid temperature changes ar e not expecte d an d
to period s durin g th e da y when superstructur e movements ar e smallest ,
generally shortl y afte r th e ambien t temperatur e approaches, reaches, and
departs fro m th e day' s pea k temperature . Backfil l a t abutment s mus t b e
placed an d compacte d simultaneousl y a t both abutment s t o avoi d displacin g
the superstructure.
6 Summar y
The firs t bridg e wit h semi-integra l characteristic s simila r t o thos e tha t have
been describe d i n thi s pape r was constructe d in 1978 . Thi s bridg e service s
Ohio's Rout e 55 5 and span s th e Muskingu m Rive r at Zanesville, Ohio . I t is a
540 foo t (164. 6 m ) lon g unskewed thre e span girde r structure. I t employ s
approach slab s an d turn-bac k wingwalls t o engag e o r embrace the backfill .
Since the n a numbe r o f simila r shorte r structures have bee n constructed . Th e
concept ha s bee n use d mos t ofte n t o retrofi t existing end-jointe d bridges .
Presently, a numbe r o f othe r semi-integral bridge s ar e being planned fo r bot h
new an d retrofi t applications , som e wit h significan t skews . A s o f thi s writing ,
skewed semi-integra l bridge s wit h guid e bearing s hav e no t yet bee n
constructed.
The respons e o f loca l maintenanc e engineer s t o thes e bridge s ha s bee n
good. I t i s primaril y through thei r continua l urgin g tha t many o f thes e bridge s
were built .
For ne w structures , the mai n emphasi s o f th e Ohi o Departmen t o f
Transportation (ODOT ) i s o n th e constructio n of fully integra l bridges with
flexible abutments . However , fo r thos e application s wher e rigi d abutment s are
necessary, th e semi-integra l bridg e concep t i s no w bein g adapte d an d use d
with increasin g regularity . Th e actua l performance of thes e bridge s
throughout th e nex t severa l year s will influenc e it s furthe r development an d
ultimately it s suitabilit y fo r furthe r applications.
7 Reference s
Burke, M . P. , Jr. , (1987 ) Bridg e Approach Pavements, Integral Bridges, and

Cycle Contro l Joints, Transportation Research Record 1113,
Transportation Researc h Board, Nationa l Researc h Council, Washington ,
DC.
NAVFAC, (1982 ) Foundation s an d Eart h Structures , Design Manua l 7.2,
Department o f th e Navy , Nava l Facilitie s Engineerin g Command,
Alexandria, VA .
R E P O R T O N SESSIO N SI X
D.W. QUINIO N
The titl e 'retrofitting ' was reall y a misnomer for two papers on experienc e in Japa n
and the USA. M r Takano amplified o n the experience gained in Japan with various
types o f embedded-type jointing s an d combined-type jointless methods . Th e former
are extensivel y use d o n th e basi s o f predicte d performanc e an d th e accumulatin g
experience of test programmes . Th e joint variants allow fo r the differen t pavemen t
characteristics. Th e jointless metho d i s a treatment providing structura l continuity .
It was commente d tha t early experience i n th e UK of suc h embedded-typ e joint s i s
very simila r t o tha t described . Ther e ha d bee n weldin g difficultie s wit h stee l
laminated bearing s bot h i n Japa n an d th e UK .
Mr Burk e describe d ho w Ohi o hav e a lon g experienc e o f buildin g semi -
integral bridges an d this type of bridge is now being used frequentl y i n neighbourin g
states learning from that experience. H e asserte d that elastomeric bearings are now
regarded a s very reliable an d no caus e for concern. H e describe d the usual ways i n
which the y ar e installe d whic h indicate d tha t the y coul d b e replace d i f necessary .
There was discussio n abou t th e sealin g o f the ga p between abutmen t an d the deck .
Suggestions wer e made that an alternative way using an access gallery could provide
easy acces s t o th e bearin g an d th e undersid e o f th e joint a t pavemen t leve l whilst
providing protectio n to the bearing against loose material s and water. Whils t it drew
on a common Germa n practice regarding access galleries M r Burke considered it to
be unnecessar y a s well a s expensive .
Mr Burk e describe d 'run-on-slabs ' a s bein g ver y reliabl e a s the y wer e
conservatively reinforced against any loading conditions . The y had never been calle d
into question . Ther e was als o discussio n o n th e needs fo r and provision o f restraint
against horizonta l rotatio n o f strongl y skewe d bridg e decks . M r Burk e ha d
emphasized th e nee d fo r guide bearing s t o prevent this . U K members commente d
that post-tensionin g o f ske w bridge deck s bearin g o n slende r pier s ca n resul t i n a
redistribution of stress and deflections unles s th e deck is restrained against rotation.
PART SEVE N
PRECAST BEA M
DECK C O N T I N U I T Y
20 C O N T I N U I T Y O F PRECAS T
PRETENSIONED CONCRET E
B E A M S B Y EMBEDMENT
IN I N T E G R A L
CROSSHEADS
B.P. PRITCHAR D
Consultant to W.S. Atkins Consultants Ltd and
Colebrand Ltd, Hay wards Heath, UK
Abstract
The relativ e advantage s o f th e fiv e type s o f continuit y method s whic h ca n b e
used fo r precast , pretensioned bridg e beam s ar e discussed. Thes e are : wid e
and narrow in-situ integra l crossheads ; integral crosshea d cas t in tw o stages ;
continuous separate d deck slab ; an d tie d dec k slab . I t then describe s a surve y o f
existing bridge s constructe d by thes e method s whic h wa s carrie d out fo r th e
Transport Researc h Laboratory.
Keywords: Precas t pretensioned concret e beams, bea m continuit y b y
embedment, dec k sla b continuity , dec k sla b hinging .
1 Introductio n
The advantage s o f precas t concrete beam dec k constructio n of eas e an d spee d

of erection , eliminatio n o f obstructiv e falsework, standardisatio n an d factory
qualities brough t t o sit e ar e well known . A s a resul t i t i s on e o f th e mos t
favoured constructio n techniques fo r smal l t o mediu m spa n bridge decks .
The mos t commonl y use d prestresse d beams ar e the 'M ' and 'L P series,
supporting an d composit e wit h a n in-situ reinforce d dec k slab . Multi-spa n
bridges rarel y feature continuou s deck s an d ar e usually buil t a s a sequenc e o f
simply supporte d spans . Wit h bearings require d under eac h beam, suppor t pier s
are o f dec k width, forme d eithe r a s soli d leaf o r columns an d crosshea d
arrangements.
Precast concret e beam dec k constructio n has generall y bee n limite d t o
simply supporte d structures . Whe n multi-span , i t i s easie r t o desig n an d
construct th e dec k a s a sequenc e o f simpl y supporte d span s tha n t o incu r the
added complexit y an d tim e penaltie s b y addin g continuity . Thes e difficultie s ar e
often mor e perceive d tha n real.
In addition , multi-spa n simpl y supporte d deck s ar e usually confine d t o
straight o r near straight alignments becaus e o f difficultie s i n fittin g straigh t beam
elements int o curve d decks. Pla n curvatur e requires varying beam length s acros s
the widt h o f eac h span , removin g mos t o f th e benefit s o f standardisation . Tigh t
vertical curvatur e leads t o excessiv e in-sit u dec k sla b build u p a t midspan o f th e
chorded beams, addin g unnecessar y dec k dept h an d weight. A s a result , thi s

230 Pritchard
form o f constructio n has rarel y been considere d fo r th e convolute d i n pla n an d

vertical profil e flyovers ofte n foun d i n tightl y confine d urba n situations.
However, th e emphasi s i s changin g an d continuit y i s no w a highly
desirable featur e because o f th e reductio n of expansio n joints. Corrosio n caused
by winter de-icin g salt s i s currentl y the mos t damagin g threa t to th e integrit y of
U K bridges . Investigation s fo r th e U K Departmen t o f Transport hav e
1
confirmed tha t most bridg e dec k expansio n joints lea k an d contribut e more than
any othe r bridge elemen t t o dec k an d substructur e corrosion damage. Th e joints
themselves als o hav e a poo r maintenanc e record .
2 Precas t concrete beam dec k continuit y
Precast concret e decks invariabl y carr y an d ac t compositely wit h a n in-sit u

concrete dec k slab . Wh y not us e som e o f th e in-sit u concret e construction in th e
pier area , wher e i t i s well suite d t o solvin g the problem s o f curvatur e an d
continuity beside s offerin g furthe r advantages in pie r constructio n and dec k
drainage arrangements ? Confinin g thi s constructio n to a n are a ove r the pie r
foundations give s readily availabl e suppor t fo r staging , wit h end embedde d
precast beam/in-sit u sla b constructio n in th e dec k area s between an d provide s
the individua l advantage s o f in-sit u an d precast construction at the localitie s bes t
suited.
The Departmen t o f Transpor t i s shortl y to issue a n Advice Not e whic h
will encourage the us e o f continuit y i n multi-spa n bridg e deck s i n orde r to
achieve th e desirabl e reductio n o f expansio n joints. Wit h the precas t concrete
beam dec k representin g on e o f th e mos t commonl y use d form s o f dec k
construction, th e Transpor t Research Laboratory recentl y commissioned a n
investigation o f th e way s i n which continuit y ca n be provide d fo r composit e
bridge deck s wher e th e primar y beams ar e of pretensione d precas t concrete . 2
The stud y identifie d fiv e distinc t method s o f building in various degree s o f

continuity an d eliminatin g intermediat e dec k expansio n joints i n th e UK . Som e
of th e method s ar e used internationally . On e o f th e method s als o overcome s
most o f th e urba n flyover curvatur e problems describe d i n th e Introduction.
The classificatio n o f th e method s indicate d tha t two mak e onl y th e dec k
slab continuou s o r hinged ove r the pier s whilst the othe r three provide ful l
monolithic continuit y connectio n o f th e beams an d sla b ove r the piers .
All method s requir e in-situ concret e transverse beams a t or near the
piers. Thes e beam s ca n be arrange d as twin, o n eithe r side o f th e pier , o r singly,
located ove r th e pier . Transvers e beam dept h ca n be les s than, equa l to , or
greater tha n th e longitudina l bea m depth . Fo r clarity, single transvers e beams
located ove r th e pie r an d pier bearings ar e designated a s integral crossheads.
All othe r transverse beams ar e designated a s diaphragms.
2.1 Typ e 1 - Wide in-situ integra l crosshead

This typ e o f continuit y detai l (Fig . 1 ) uses precas t pretensioned beam s
significantly shorte r than the span s betwee n suppor t pier s an d is th e mos t
Precast concrete beams in integral crossheads 23 1
Beam Beam
Top slab/crosshead
Top o f bea m continuity continuity
reinforcement \ . reinforcement >
3s
4
,...:.|V.
i
Crosshead soffi t m a y ' , Transverse
* reinforcement or
be flus h with beam
soffit | j — —jl prestressing through
Bottom continuit y I beam web holes
Beam embedment reinforcement, '
Temporary trestl e
'1
supported o n
Crosshead lengt h
pier foundation
Fig. 1 Typ e 1 continuity connection : wid e in-sit u integra l crosshead .
commonly use d U K method . Th e beams , generall y 'U ' o r 'M ' type, ar e usuall y
supported o n temporar y trestles buil t of f th e pie r foundations. Th e wide in-sit u
integral crosshea d ove r th e pie r i s the n cas t between an d aroun d the beam s t o
provide abou t 1 m embedment . Longitudina l continuit y i s accomplishe d b y
reinforcement withi n the continuou s composit e dec k slab , generall y
supplemented b y reinforcemen t and , sometimes , pretensionin g stran d end s
extending fro m th e to p an d bottom o f th e embedde d beams . Transvers e
strength i s assure d b y eithe r prestressin g tendon s o r reinforcement, som e o f
which ma y pas s throug h hole s i n th e end s o f th e precas t beams. Th e crosshea d
is supporte d o n a singl e ro w o f bearings se t centrall y on th e pier .
2.2 Typ e 2 - Narrow in-situ integra l crosshea d

This continuit y detai l i s use d extensivel y i n Americ a (Fig . 2) . Th e precast
pretensioned beam s ar e long enoug h t o b e erecte d o n tw o paralle l row s o f
temporary o r permanent bearing s o n th e pie r tops . A s wit h Type 1 , the in-sit u
integral crosshea d ove r th e pie r i s the n cas t between an d aroun d the beam s t o
provide abou t 1 m embedment . Th e crosshea d is , however , narrowe r than
Type 1 because o f th e necessaril y smal l ga p between th e embedde d beams . Th e
same narro w gap make s adequat e reinforcemen t connectio n difficul t betwee n
the bea m ends . Longitudina l hoggin g bending continuit y i s agai n readil y
established b y to p reinforcemen t withi n an d extendin g wel l into th e continuou s
composite dec k slab . Transvers e strength o f th e crosshea d i s generall y provide d
for b y reinforcement , som e o f which passe s throug h holes in th e end s o f th e
precast beams .
Where twi n rows o f temporar y bearings ar e used, a centra l row o f
permanent bearing s locate d unde r th e crosshea d i s brough t int o us e b y removin g
the temporar y bearings afte r th e crosshea d concret e has gaine d sufficien t
232 Pritchard
pier
Beam embedmen t T o p s lab/crosshead

~~| T continuit y reinforcemen t
4
Precast b e a m N
Precast bea m
Transverse
reinforcement Bottom continuity
through bea m reinforcement
webs
Crosshead
length
Fig. 2 Typ e 2 continuit y connection : narro w in-situ integra l crosshead.
strength. Som e example s us e a wide singl e permanent rubber bearing whic h

acts a s a seatin g fo r both beams .
2.3 Typ e 3 - Integral crosshead cas t in tw o stage s

This typ e o f continuit y detai l i s a variant of Type s 1 or 2 where th e integral
crosshead i s cas t in tw o stage s (Fig . 3). Th e crosshea d is o f greate r dept h tha n
the mai n dec k precas t beams an d th e botto m sectio n i s cas t first t o suppor t
these beams , generall y o n thi n morta r beddings . Th e secon d stag e complete s
the integra l crosshea d in th e manne r described fo r Typ e 1 .
2.4 Typ e 4 - Continuou s separate d deck sla b

This continuit y detai l confine s itsel f t o th e dec k sla b only , whic h flexe s t o
accommodate th e rotation s of th e simpl y supporte d dec k beams erecte d in th e
conventional multi-spa n manne r (Fig. 4). T o permi t this flexure , th e dec k sla b i s
separated fro m th e suppor t beams fo r a length o f abou t 1. 5 m by a layer of
compressible material . In-sit u reinforce d concret e diaphragms are located a t the
ends o f th e separate d slab.
A variatio n of thi s metho d omit s th e diaphragm s and is know n a s th e
continuity plat e construction . Thi s was introduce d in Americ a in th e fiftie s an d
3
tested an d develope d i n Switzerlan d an d Germany in th e sixties . I n Bavaria ,

Germany, ove r 30 0 multi-span precas t concrete beam bridges ha d been buil t b y
1981 usin g this Typ e 4 variant. Th e detai l was reporte d as practical, economical
and durabl e an d is stil l being constructe d in Germany.
2.5 Typ e 5 - Tied dec k sla b

This tie d dec k sla b detai l wa s develope d fo r multi-span precas t beam deck s
during th e U K Standar d Bridge exercise o f th e 1970s . Th e bridge deck s are
4
Precast concrete beams in integral crossheads 23 3
Crosshead Extent o t stag e l

Top slab/crosshea d
construction
continuity reinforcemen t
depends o n
Beam,
Mortar be d transverse
or pa d bearin g N
'design
Top o f bea m - . Bottom continuit y

J-r continuit y reinforcement
reinforcement Precast bea m
In sit u
Precast bea m
crosshead,
Crosshead- second stag e
monolithic wit h pie r
In sit u crosshead ,
or o n bearin g wit h
first stag e
temporary fixit y
during
construction
Reinforcement omitte d fo r clarit y
Fig. 3 Typ e 3 continuity connection : integra l crosshea d cas t i n tw o stages .
Separated dec k
slab lengt h Sla b reinforcemen t
Compressible fille r y _ continuous acros s
between bea m an d joint betwee n beam s
slab \
\ ~ 4
Precast bea m Precast bea m v
1 1 Transverse
reinforcement
through bea m
webs
Fig. 4 Typ e 4 continuity connection : continuou s separate d dec k slab .
designed an d constructe d i n th e conventiona l multi-span simply supporte d

manner wit h slab trimme r diaphragm s a t th e bea m ends . Thes e end s are , a s
with Type 4, carrie d o n tw o paralle l row s o f bearing s o n th e piers .
Long connectin g reinforcemen t dowel s ar e incorporate d a t th e sla b
mid-depth t o ti e th e slab s togethe r ove r th e pier , eliminatin g expansio n
movement a t dec k leve l an d permittin g th e us e o f burie d dec k rotatio n joints .
To accommodat e thi s rotation, th e dowel s ar e debonde d an d sleeve d fro m th e
surrounding sla b concret e ove r shor t length s eithe r sid e o f th e joint . Also , th e
slab an d trimme r bea m downstand s ar e 'necked ' using compressible joint fille r
below an d abov e th e dowe l connection .
234 Pritchard
Crack-inducer slo t
in surfacin g filled
with rubber/bitumen
seal
Galvanized hig h
yield strength
dowel ba r
Precast bea m
Transverse
reinforcement
through bea m webs
Fig. 5 Typ e 5 continuit y connection : tie d dec k slab .
3 Choic e o f continuity metho d
If th e mai n requiremen t is t o eliminat e troublesom e intermediat e dec k

expansion joint s i n multi-spa n deck s the n Type s 4 an d 5 offe r th e minimu m o f
extra desig n an d constructio n effort. Th e simples t procedure , Type 5 , is
however probabl y the leas t durabl e because th e tie d dec k joint retain s a
rotational capacity , with a seale d notc h require d in th e dec k surfacing .
However, t o gai n th e advantage s o f par t dec k continuity , on e o f th e Typ e
1, 2 o r 3 methods shoul d b e adopted . Th e dec k bending reduction s ar e modest
but stil l worthwhile i n term s of constructio n depth saving s an d associate d saving s
in th e length s o f flyover s and/o r approach ramps. Modest , becaus e th e precast
beam self-weigh t elemen t remain s simpl y supporte d an d further continuit y
moments du e t o restraine d creep and shrinkag e o f th e beams mus t b e
considered i n additio n t o th e norma l surfacing, liv e load an d temperature
loadings. Nevertheless , typica l saving s allo w M 8 beams t o be use d i n plac e o f
simply supporte d M1 0 beam decks .
The embedmen t o f th e precas t pretensioned beam s int o th e in-sit u
crosshead involves restrain t to th e continue d prestressin g cree p rotation of th e
beam ends , givin g rise t o saggin g moment s a t the supports . Th e beam end s
rotate i n th e opposin g directio n du e t o th e late r differential shrinkag e se t u p i n
pouring th e in-sit u concret e deck slabs , givin g rise t o hogging moments ove r th e
supports. I n general , th e tw o moment s ten d t o eventuall y cance l eac h other.
However, ther e is usuall y a on e o r two mont h dela y between bea m embedmen t
and sla b pourin g an d th e moment s du e t o th e restraine d creep during tha t dela y
period mus t b e considere d alon e fo r on e o f th e desig n loading combinations .
In the followin g sections , th e relativ e advantage s o f Type s 1 and 2
Precast concrete beams in integral crossheads 235
methods ar e listed. Typ e 3 method , representin g a n amalga m o f Type s 1 and 2 ,

is no t particularl y favoured fo r tw o reasons . Firstly , the downstan d hal f o f th e
crosshead i s obstructive , both aestheticall y an d in term s o f headroom . Secondly ,
to tak e advantag e o f a singl e row o f bearing s unde r th e crosshead , specia l car e
must b e take n durin g bea m erectio n t o preven t crosshea d rotatio n an d collapse ,
which ha s happened , du e t o unsymmetrica l erectio n o f th e precas t beams o n
each sid e o f th e crosshead .
3.1 Th e relativ e merit s o f Type 1 metho d

This, th e mos t commonl y use d metho d i n th e UK , i s relativel y mor e comple x t o
design an d construc t than an y o f th e fou r othe r methods. Nevertheless , i t offer s
more advantages , includin g th e abilit y t o accommodat e plan an d vertical
curvature.
The ke y elemen t i s th e wide in-sit u concret e integral crosshead , with
temporary beam suppor t trestlin g an d crosshea d formwor k supporte d of f th e
pier foundations . Th e wide crosshea d offer s th e followin g advantages :
(a) Precas t bea m dec k span s ar e most ofte n limite d t o th e readil y

transportable lengt h o f 27. 5 m define d b y th e Departmen t o f Transport .
With 5 m wide crosshead s an d 1 m beam embedments , th e continuou s
span length s ca n be extende d t o 30. 5 m. Th e usua l en d projectin g
reinforcement ca n be remove d fro m embedde d scre w coupler s durin g
transport.
(b) Pla n curvatur e ca n be readil y accommodate d by varying th e widt h o f th e
integral crosshea d t o for m a pla n trapeziu m shape . Thi s overcome s th e
usual proble m o f th e beam s o n th e outsid e o f th e curv e being longer tha n
those o n th e insid e an d permit s th e us e o f on e standar d length beam .
The sid e curvatur e i s take n u p b y varying th e widt h o f th e facin g
concrete.
(c) Vertica l curvatur e problem s ca n be reduce d b y vertically curvin g th e to p
and bottom surface s o f th e crosshead . Thi s reduce s th e increase d sla b
thickness a t midspa n require d t o tak e u p th e vertica l curvature abov e th e
straight chorde d precas t beams.
(d) Drainag e o f surfac e water o n multi-spa n viaduct s ideall y require s
collector downpipe s space d n o furthe r apar t tha n th e pie r suppor t
positions. Thi s mean s latera l connecting pipin g between ker b gulley
drains a t th e dec k edge s an d downpipe s within o r attached to th e side s of
the piers . Norma l simpl y supporte d composit e bea m spa n deck s ar e so
congested wit h precast beam end s ove r th e pier s tha t it i s usuall y to o
difficult t o pas s th e drainag e pipes through . Resor t has therefor e t o b e
made t o ugl y externa l piping passing ver y prominently dow n the dec k
sides o r to mor e comple x longitudina l dec k drainag e systems . Th e
Type 1 continuity metho d uses wide in-sit u integra l crosshead s wit h the
precast bea m end s embedde d som e distanc e fro m th e pier . Thi s readil y
allows enoug h roo m fo r th e passag e o f a conceale d interna l lateral
connecting drai n using generous diameter s an d bends (Fig . 6).
236 Pritchard
i
180 mm thick slab
between beams
150 mm thick slab Two 120 mm x 30 mm
over beams | ducts per tendon Ten 0-76 Strandforce
anchorages
Substantial embedded
drains from kerb gullies
to concealed slot in pier
column
Fig. 6 Conceale d drainag e piping.
(e) Typ e 1 uses a singl e row o f bearings. Thi s immediatel y halve s th e

numbers o f bearing s require d for simpl y supporte d construction , althoug h
individual bearin g siz e will marginally increase.
It als o mean s tha t pier s ar e thinner, no t onl y becaus e th e singl e line o f central
bearings take s u p les s room a t the pie r top , but becaus e th e dea d an d live loa d
moments applie d t o th e pier s b y of f centr e pairs o f bearings ar e removed.
Significant saving s i n pie r foundation s als o result .
Full width pier s o r crosshead s ar e not require d either . Th e integra l
crosshead i n th e dec k ca n be to p reinforce d o r prestressed laterall y to allo w
considerable dec k cantileverin g outsid e th e pier . Resultin g inboar d pier s ca n
offer considerabl e saving s i n groun d intrusion . Thi s i n tur n can allo w groun d
level sli p roads t o b e locate d unde r th e dec k edge s an d ca n reduce flyove r span s
at ske w crossings , bot h particularl y important in crampe d urban locations .
There i s als o a useful savin g i n furthe r numbers o f bearings .
The larges t us e o f Typ e 1 method i n th e U K was durin g th e constructio n
of th e si x viaduct Londo n t o Cambridg e M il Sout h Woodfor d Interchang e in
1976 . Thi s wa s followe d b y th e constructio n in tw o furthe r similar viaduct s
5
forming par t o f th e Sout h Woodfor d t o Barkin g Relief Roa d in 1987 . Th e

crossheads wer e transversel y prestressed , a s much a s anythin g t o avoi d
congestion i n th e uppe r surfac e du e t o heav y transvers e and longitudina l
reinforcement. Test s o f th e crosshea d were undertake n a t the the n Cemen t an d
Concrete Associatio n a t Wexha m Springs . Thes e demonstrate d th e
6
effectiveness o f th e shea r transfer between th e beam s an d crosshea d an d prove d

the metho d t o b e entirel y successful . Nevertheless , th e transvers e prestressing i s
expensive an d late r constructions hav e resorte d to th e us e o f norma l stee l
reinforcement rathe r tha n prestressing .
3.2 Th e relativ e merit s o f Type 2 metho d

This i s almos t th e onl y continuit y metho d use d i n Americ a and is relativel y
easier t o build tha n Typ e 1 . Nevertheless , i t canno t offe r th e advantage s o f
extra span , curvatur e o r drainage offere d b y Typ e 1 . I n most cases , i t als o doe s
not offe r th e bearin g an d pie r advantages , unles s beams ar e seated upo n
Precast concrete beams in integral crossheads 2 3 7
temporary twi n rows o f bearings an d loads ar e eventually transferre d to a

permanent singl e ro w o f bearings se t unde r th e integra l crosshead , o r use i s
made o f a wide singl e bearing .
Undoubtedly, th e greates t advantage s o f Typ e 2 method lie s in th e eas e
of placin g th e precas t beams directl y o n t o th e piers . Nevertheless , th e narrower
integral crosshea d make s adequat e reinforcemen t mor e difficul t betwee n bea m
ends, particularl y for saggin g momen t botto m reinforcement .
A recen t American report has propose d tha t the provisio n o f thi s botto m
7
reinforcement shoul d b e optiona l provide d tha t continuit y moment s ar e

calculated assumin g lac k o f positiv e momen t continuity . I t is als o suggeste d tha t
any crackin g resulting fro m no t providin g thi s reinforcemen t i s n o threa t to th e
deck integrit y an d will certainl y eliminate an y corrosio n possibility. N o doub t a
compromise solutio n coul d b e use d - a limited amoun t o f anti-crack
reinforcement i n th e botto m o f th e crosshead , possibl y epoxy coate d t o preven t
corrosion.
Major researc h and testin g o f th e Typ e 2 connectio n wa s undertake n b y
the Portlan d Cement Associatio n i n Chicag o in th e earl y sixties . I t covere d al l
8
aspects, rangin g fro m cree p an d shrinkag e t o horizonta l shea r connection . I t

also covere d extensiv e testin g o f a half scal e continuou s tw o spa n deck , som e 2 0
m lon g x 5 m wide . A recen t UK publication set s ou t desig n procedures for
9
using Type 2 continuit y fo r th e ne w precas t Y beams .
4 Reference s
1 Wallbank , E . J. (1989) Th e performance of concrete in bridge s - a surve y

of 20 0 highway bridges . Londo n HMSO .
2 Pritchard , B . P. an d Smith , A . J . (1991) Investigatio n o f methods o f
achieving continuit y i n composit e concret e bridge decks . T R L
Contractor Repor t C R 247 . Crowthorne .
3 Wagne r and Buchtin g (1981 ) Continuit y plat e constructio n i n multi -
span bridge s usin g precas t concrete beams. Bauingenieur .
4 Departmen t o f Transpor t (1979) DT p standar d bridges. Londo n HMSO .
5 Pritchard , B . P. (1982 ) Integra l in-situ crosshead s fo r continuou s precas t
beam dec k viaducts . Internationa l Conference o n shor t an d mediu m
span bridges . Toronto .
6 Sturrock , R. D . (1974 ) Test s o n mode l bridg e bea m i n in-sit u concret e
construction. Cemen t an d Concret e Association .
7 Transportatio n Research Board (1989) Desig n of precast prestressed
bridge girder s mad e continuous . Nationa l Cooperativ e Highwa y
Research Progra m Report 332.
8 Mattock , A. H . Kaar , P . H. Hanson , N . W . Hognestead, E . and Kriz ,
L. B . (1960-61 ) Precas t concrete bridges. Seve n paper s i n Journa l o f
Research an d Developmen t Laboratories . Portlan d Cement Association .
9 Hambly , E . C. an d Nicholson , B . A. (1991 ) Prestresse d beam integra l
bridges. Leicester , Prestressed Concret e Association .
21 L O C A L L Y SEPARATE D
DECK SLA B CONTINUIT Y
IN COMPOSIT E BRIDGE S
A. KUMA R
Kumar Associates, Beaconsfield , U K
Abstract
This paper considers the problems associated with jointed deck construction and com-
monly used remedies leading t o logical evolution o f ideas and concepts of locally separat-
ed deck slab continuity in composite bridges. These are further advanced to practically
usable details in harmony with the overall structural behaviour of bridges. Proper applica-
tion of these should result in savings to the exchequer, increased life and trouble free
performance of such bridges.
Keywords: Dec k slab continuity, Top slab continuity, Locally separated deck slab,
Connecting slab , Continuity plate, Isolation plate, Multi-span composite bridges, Contin-
uous composite bridges, Continuity with precast bridge beams, Continuity with steel
beams, Concepts of continuity.
1 Backgroun d
Span by span simply supporte d construction of multi-span bridges using some form of
precast (or steel) longitudinal bea m elements ha s been in vogue sinc e about 1950's.
During the sixties and seventies the need for rapid expansion of the country's trunk road
network combined with no falsework requirement and shorter period of deck construc-
tion, the numbers of such multi span bridges rapidly increased. Due to the public require-
ment of keeping roads open through the winter months, this period also witnessed an
increasing use of salt on British roads.
During the seventies inspections o f such bridges had started revealing that despite every
effort in devising competent joint details, salt water penetration through the joints and
consequential damag e to deck ends and support elements, wa s wide spread. The pres-
sures of rapid expansion of roads network and simplicity o f design and construction are
however such that the practice of jointed construction has continued largely unchallenge d
through the eighties to the present day.

240 Kumar
Estimating the cost implications of deterioration due to such construction is quite difficul t
as it involves varying combinations of repair and replacement costs, reduction in useful
life of structures and traffic disruption etc. Even if the latter is excluded, preliminary
estimates place the one going cost of such deterioration at around £10m/year. Althoug h
this is not a particularly large sum in the context of government's expenditure on roads, it
is the author's view that such direct costs together with associated unquantifiable traffi c
disruption costs, would be significant an d could be largely eliminated by using the con-
cepts presented here.
2 Existin g remedie s
As mentioned earlie r the problem of joints deterioration has been recognised for some 20
years in the UK and for about 30 years in the USA. During the early sixties in the USA
and seventies i n the UK, researches had been conducted to eliminating joints in such
construction. The resulting so-calle d live load continuity solutions i n which large insitu
1
concrete diaphragms embedding th e ends of precast units are used, do eliminate the
intermediate support joints and hence such deterioration. However the design process is
greatly complicated due to secondary effects suc h as temperature, shrinkage and creep
becoming importan t design parameters. The manufacture of precast elements an d con-
struction of the deck are also made considerably more complicated because of the need to
accommodate reversals of moments in the precast and insitu concrete adjoining the
intermediate support positions. Thi s also generally negates any anticipated reduction in
depth of construction of such decks. These factors in reality can lead to cost increases of
the order of 25% to 50% in the cost of deck construction depending o n the number of
spans involved. Suc h increases in cost could amount to additional expenditure of up to
£10m per year, if this remedy is applied to all such bridges.
The so called 'integral' bridges could also be viewed as an extension of this approach, in
that the abutments are also integrally connected to the bridge deck with no provision fo r
movements except by sliding of abutments on foundations. I t is however doubtful wheth -
er the additional expense and empiricism in design is justifiable agains t the cost saving of
an end expansion joint as well as the likelihood of disintegration of soil wedges behind
abutments and indeterminate state of stress in the structure.
To avoid additional costs of such order a relatively simple method in which simpl y
supported decks are tied together by locally debonded 'tying ' bars at deck ends, has been
used in some bridges during the mid eighties. Result s of their performance are not fully
know yet. However, complicated detailing an d possible poor workmanship, the presence
of a shrinkage crack at the throat and concentrated rotations at deck ends affectin g
surfacing, ar e potential causes for concern in the long term.
3 Propose d approach
The locally separate d deck slab continuity concept as presented here avoids complicate d
Deck slab con tin uity in composite bridges 241
design and construction and may even result in some savings below conventional jointed
construction due to simplifications i n detailing an d drainage arrangements at these loca-
tions.
In view of the durability problems associated with the conventional jointed construction
and great complexity of design and construction of live load' continuity decks, the author
invented thi s concept about 10 years ago. The method is briefly describe d in references 1
and 2 and has now been implemented i n some eight bridges in the UK including th e 5
span 110 m long river Frome bridge, with Department of Transport approval. It is no w
know tha t several other countries notably Germany have used similar ideas on large
numbers of bridges apparently with considerable success . It is however a sad commentary
on European Community co-operation that important technical matters of day to day
concern to engineers ar e not being communicated across linguistic boundaries .
4 Th e conceptual basi s
In typical single span precast beam bridge construction, a relatively thin top slab is cus-
tomarily used to affect load distribution between th e beams. Depending on load position -
ing, the beam elements rotate and deflect differentiall y t o each other causing the deck slab
to flex to accommodate such displacements. Th e structural action of the slab is complex
and involves interactions of flexure, shear , torsion and some local effects. Thi s load
distributive action of the slab can be viewed as two simple modes of flexural actions viz
(a) symmetric and (b) asymmetric bending of a pin ended slab element a s depicted in Fig.
1. It will be noted that mode (b) here allows the relative deflections o f the ends in relation
to the deflected en d tangents. Various deflected form s of a slab element du e to end
deflections o r rotations are linear combinations of these two modes. In additio n there is
also the equilibrium actio n of the slab when loads are applied within its span.
The ends of bridge decks simply supporte d on metal or elastomeric bearings at interme-
diate supports undergo small rotations and deflection relativ e to each other depending on
load positioning on the deck. Although th e main spans of the bridge are large in compari-
son to slab spans between beams , the depth of construction and stiffnesses are large too,
(a) Symmetric mode (b) Asymmetric mode
Fig. 1 Modes of flexure of transverse slab

242 Kumar
resulting in similar order of rotations as for the slab. It therefore follows that slab panels
of similar proportions as are conventionally used between the precast beams, should be
able to accommodate movements at span ends, a matter to be confirmed by the detailed
design process. It will be noted that this deck slab continuity by itself, will not cause any
net forces in the connecting slab panels. Fig. 2indicates the flexing of such a slab in the
basic and combined modes.
(a) Symmetric rotation (b ) Relative deflection an d rotation
Fig. 2Mode of flexure of connecting slab panel at support
Under design loadings the main spans will deflect in the sagging mode and the connect-
ing slabs in hogging mode. In the absence of live loads but due to temperature difference
and creep effects th e span elements could be hogging upwards and the connecting slabs
sagging downwards. These two extreme conditions are indicated in Fig. 3. Connecting
slabs will thus be flexing in reverse curvature to the main span members of the bridge. It
will be noted that decks with such continuity of displacements between spans will provide
a suitable base for laying waterproofing and surfacing for vehicular use.
(a) spans sagging
(b) Spans hogging
Fig. 3 Longitudinal action of connecting slab panels

Deck slab con tin uity in composite bridges 24 3
5 Applicatio n
For practical applications of this concept to composite bridges the following aspects need
to be considered:
(a) I n order to accommodate connecting sla b panels between span s without widening
the supports, it would in general be necessary to locally separat e the deck slab from the
tops of beams, thus creating independently actin g slab elements a s shown in Fig. 4. The
width of supports required for such arrangement would be the same as for simply sup -
ported construction.
(b) Transvers e diaphragms should be provided set at connecting sla b edges a s shown in
Fig. 4. This will minimise punching shears , allow easy jacking for probable bearing re-
placement, reduce relative displacements o f adjoining beams and contribute to load distri-
bution properties of the deck.
(c) Connectin g sla b may also be used at one of the ballast walls as shown in Fig. 5 and
the entire longitudinal movemen t of the deck accommodated at the other abutment. For
this option connecting sla b should preferably be located at downstream end to allow
shedding of water from the bridge deck without percolation through the joint. Howeve r
locating connecting sla b here, would reduce ability for jacking for bearing replacement
therefore high stiffness bearings such as metal bearings requiring only smal l amount of
lifting would be more appropriate here. For obvious reasons , connecting slab s at both
abutments are not recommended for multi-span bridges.
(d) A t significant parape t upstands at intermediate support positions, th e upstands

should be jointed and separated fromthe connecting sla b as shown in Fig. 6. If however
Compressible separating
layer.
Fig. 4 Locally separated deck slab with transverse diaphragm

244 Kumar
Fig. 5 Connecting slab also at one of the abutments
Fig. 6 Separating and jointing o f parapet upstand
these upstands are not too deep or significant cantilever s or wide footpaths exist on the
deck, it may be possible to control any cracking without the need for separating and
jointing th e upstands.
6 Desig n considerations
As explained before, the mode of flexural action of the connecting slab s is opposite to the
main spans. Despite it s ability to support local wheel loads, its contribution towards
restraining the main spans against end rotations is relatively modest. It would therefore be
simple an d conservative to design the main spans without taking advantage of this re-
straint. The design of the connecting slab s can then be conservatively based on the calcu-
lated end rotations of the span members. The connecting sla b proportions are generally
such that these rotations will indicate cracking which should comply with Code require-
ments.
It therefore follows that should restraint advantage from the connecting sla b be desired
to be considered in the global analysis of the main deck, only the cracked stiffness of this
component is used in the model, irrespective of cracked, uncracked or gross inertia s
being chosen for the representation of deck members in accordance with the Code.
Deck slab con tin uity in composite bridges 24 5
The length, thickness and reinforcing of connecting slabs will normally depend on
structural configuration, range of ages of beams, level of prestress, live loads, local wheel
loads, stiffness o f bearings, temperature differences, differentia l settlements , skew and
method of articulating the bridge. This approach is equally applicable with the M, Y and
U beam bridge decks currently used in the UK. Moderate lengths of connecting slab are
likely to prove adequate as local and global effects will tend to counteract each other.
Fatigue in reinforcement due to stress reversals, should be designed 'out ' by compliance
with Code requirements, as for the rest of the deck slab.
Preliminary calculations indicate that possibly u p to 12 spans could be made continuous
using this approach. However, it should be noted that as the number of spans increases,
the forces due to friction or shear in the bearings and due to dynamic effects du e to
pulling of many spans, would tend to increase in the connecting slabs. Without detailed
calculations for these effects, no t more than about 6 spans should be made continuous
using this technique.
7 Europea n experience
As mentioned earlier it is now known that similar ideas have been applied elsewhere in
Europe notably in Bavaria in Germany and to a lesser extent in Norway. In German
practice outlined in reference 3and Norwegian practice outlined in reference 4, this
approach of slab connection is referred to as 'continuity plate' and 'isolation plate con-
1
struction respectively. Although the success largely depends on the detailed arrangements
adopted, the experience of these countries with this approach is thought to be very satis-
factory.
8 Advantage s
The principal advantage of this approach is its great simplicity of design and construction
combined with effectiveness. B y thoughtful design this approach could result in first costs
below simpl y supporte d construction due to simplification o f details of diaphragms,
joints, parapets and drainage system.
In the live load' continuity approach the cracking of the top slab adjoining the continui-
ty diaphragm member is likely to extend to the full depth of the deck slab due to global
tension arising here under live loads. In the proposed approach here, the connecting slab
predominately acts in flexure therefore any cracking would tend to be limited to one face
of this slab, the other face being in compression thus reducing the possibility o f water
penetration.
Because of flexible connection between spans, the maximum reactions likely to arise on the
support elements could be less than 'live load' continuity approach. Also this approach would
provide greater tolerance towards differential settlement s of supports.
246 Kumar
9 Conclusion s
The approach described is of wide applicability in composite concrete bridge construction and
has now been designed in 8 bridges in the UK including one with 30 degrees skew. For bridges
to be designed as simply supported in compliance with the Codes, it may be possible to develop
some standardised recommendations for the design of this element
For steel/concret e composit e bridg e constructio n wher e stee l bea m continuit y i s no t
incorporated by splicing, the concepts described in this paper can be used with span by span
construction to achieve deck slab continuity to advantage.
10 Reference s
Kumar A., Continuity in composite concrete bridge construction. International confer-

ence on short and medium span bridges, Montreal, Canada, August 1986.
Kumar A., Composite concrete bridge superstructures. Wexham Springs, British Cement
Association, 1988 . 46pp. Publications 46.505.
Wagner F J., Nuremberg and Buchting F, Munich. Die Federplattenbauweise bei
mehrfeldrigen Fertigtragerbrucken aus Beton (Continuity plate construction in multi-
span bridges using pre-cast concrete beams), Bauingenieur, Springer Verlag 1981 .
Normaler Handbok -100, Bruprosjekterin g - 08 (Nib-Bruer) and - 09 (Nob-Bruer) '83,
Statens Vegvesen, Vegdirektoratet, Oslo.
R E P O R T O N SESSIO N SEVE N
G.P. TILLY
Mr Pritchard described his review of the techniques used to make precast prestressed
beams continuous. H e favoured the use of wide i n situ cross heads because they can
accommodate dec k plan and vertical curvature and avoid the use o f half-joints. Th e
effects o f cree p an d shrinkag e balance ou t wit h time . M r Pritchard cite d th e cas e
of a bridge unde r construction in Kuwait when th e beams being place d o n th e pier
head cause d i t t o deflec t s o tha t collapse occurred . A questione r commente d that
this system of construction was considered for Killikrankie Bridge; the stability of the
pier top was investigate d an d a possible solutio n wa s to use a n overhead gantry. I n
the even t thi s procedur e was not adopted . I t was pointe d ou t that it is necessary to
consider the development o f cracking in the soffit concret e over the support. Hybri d
beams, combinin g straigh t pretensioned tendon s an d ducted post-tensioned tendon s
can be use d a s a n alternativ e to drape d pretensioned tendons .
Dr Kuma r presente d hi s pape r an d explaine d tha t wit h continuou s
connections th e resultin g saggin g movement s coul d b e th e sam e a s fo r simpl y
supported design bu t hogging movement s ar e higher. I t is possible t o eliminate th e
end join t bu t th e abutmen t ha s t o slid e o r rotat e t o accommodat e movement .
Overall it is necessary to weigh up the complexity of the design agains t the advantage
of eliminatin g expansio n joints. H e the n explained th e advantages of separating the
deck sla b ove r th e support s an d makin g tha t alon e continuous . A speake r
commented tha t i n German y i t ha d bee n commo n practic e t o hav e sla b onl y
continuity betwee n precas t beams sinc e th e 1970s . I t i s possibl e t o d o thi s mor e
economically than for simply supporte d construction because cheaper beams can be
used.
The autho r responded tha t it i s no t surprisin g tha t the techniqu e i s use d i n
Germany, th e onl y surpris e i s tha t i t ha s no t bee n use d mor e i n th e UK . I t i s
feasible t o us e shorte r beam s bu t ther e ar e difficultie s i n reducin g th e saggin g
movements. Ther e are also possibilities fo r using differen t design s o f anchorage and
shallower beams . I t is necessar y for th e dec k sla b t o be abl e t o defor m an d follow
the movement s o f th e structura l elements .
PART EIGH T
SOIL-STRUCTURE
INTERACTION
22 S O I L - S T R U C T U R E
INTERACTION:
C E N T R I F U G E MODELLIN G
OF I N T E G R A L BRIDG E
ABUTMENTS
S.M. SPRINGMA N an d A.R.M. NORRIS H
Department of Engineering, University of
Cambridge, UK
Abstract
Joint-free 'integral ' bridge s ar e no w bein g recommende d b y cost-consciou s transpor t
authorities fro m aroun d the world . Avoidin g th e problem s associate d wit h corrosio n
and mechanical disruption o f expansion joints, by excluding them , creates a situation i n
which therma l strain s induce d i n th e dec k caus e cycli c loadin g o n th e soi l behin d th e
abutments. Ther e are two aspect s of soi l behaviour which ar e of concer n to the design
engineer - whether the soil behind th e wall wil l settle significantl y an d whether there is
a buil d u p i n horizonta l stresse s whic h wil l caus e a sever e increas e i n th e bendin g
moment carried by the abutment wall.
Keywords: Soil-structur e interaction , Integra l bridges , Abutments , Temperatur e
effects, Strai n cycling, Non-linearity , Hysteresis .
1 Introductio n
The Transpor t Research Laborator y hav e commissione d a serie s o f centrifug e mode l

tests t o examine th e behaviou r o f full-heigh t 'rigi d frame' or 'integral' bridges retainin g
cohesionless soil . Th e rational e fo r bridg e desig n usin g thi s metho d i s alread y wel l
documented (Hambly , 1991 ; Pritchard, 1992) .
Two different type s o f abutment will be investigated: a relatively flexibl e piled wall
with a stiff dec k (Figur e la), or a stiffer sprea d base abutment combined wit h the same
deck (Figure lb). I n the first case, the mode of behaviour is anticipate d to be flexure of
the pile wal l combine d wit h a rigid connectio n whic h force s rotatio n a t the wall-dec k
joint. I n the secon d example , th e displacement a t deck level will be accommodate d by
a combination o f rotation, translational shear and perhaps a degree of flexure.
This pape r wil l focu s o n som e mechanism s whic h wil l affec t th e soil-structur e
interaction, th e tes t specification s an d a brie f descriptio n o f th e equipment , befor e
quoting som e preliminary results.

252 Springman and Norrish
Concrete o r steel/concret e
composite dec k Concrete o r steel/concret e
composite dec k
6m
Drained Pile ca p / rigid abutment-dec k 6m

cohesionless connection Pile ca p / rigi d abutment-dec k
Drained
in-situ soi l connection
cohesionless
or backfil l
backfill
material "Steel pile abutment material
Reinforced concret e abutmen t
•
No momen t concret e connectio n
cohesionless soil foundatio n
cohesionless soi l foundatio n
Fig. la . Pile d integral abutment Fig . lb . Sprea d base integral abutment
2 Centrifug e model testing : Bridg e abutments
Table 1 . Summar y of centrifuge model tests on soil-structure interaction of bridges
Date Research Focus Sponsor Researche r
1984-1987 Single ro w of piles/pile groups in soft clay TRR L Springman

(10 tests) overlying san d adjacent to surcharge loads (1989)
1985-1988 Spread base abutment s on stif f clay , san d Sun (1989 )
(7 tests) embankment poured in-fligh t
1992-1993 Buttonhole foundations - pile group E C Bransby
(4 tests) adjacent to surcharge load (1992, 1993)
1992- Rigid frame bridges in sand TR L Norrish (1993)
1992- Piled full-heigh t abutments , i n sof t cla y TR L Ellis (1993 )
overlying sand , san d embankment poured
in-flight
Reports:
Contractor Report 196 : Th e effect o f TRRL Springman &
surcharge loading adjacent to piles Bolton, 1990
Contractor Report 270: Geotechnica l TRRL Bolton, 1991
stress analysis for bridge abutment design
Centrifugal modelling of abutments 25 3
The Cambridg e Geotechnical 1 0 m bea m centrifug e ha s bee n use d t o investigat e th e

behaviour of several types of bridge abutments and piers (Springman, 1989 ; Sun , 1989)
leading t o the delivery o f design recommendations (Springman & Bolton, 1990 ; Bolton,
1991). Bransb y (1992) , Elli s (1993 ) an d Norris h (1993 ) ar e currentl y engage d i n
further research on the behaviour of full-height bridg e abutments.
The advantages of modelling th e soil-structure interaction at the correct stress-strain
levels, an d a t smal l scal e i n th e centrifuge , ar e many . Th e abilit y t o observ e th e
deformation mechanism s a t serviceability limi t state , an d the n th e failuremode s a t
ultimate limi t state , is particularl y valuable. Th e tim e scale s an d financial outla y are
also significantl y les s than might b e expected for an equivalent rang e of prototyp e full
scale field tests. Schofiel d (1980 ) describes the principles of centrifuge modelling .
3 Behaviou r of integral bridge s
Diurnal an d seasonal temperatur e variations will caus e expansion an d contraction of a

continuous integra l bridg e deck , which , i n th e absenc e o f a n expansion joint , wil l b e
accommodated by cyclic lateral displacements o f th e wall-deck joint . A s th e abutment
moves awa y fro m an d toward s th e complian t fill , th e cohesionles s soi l particle s will
tend t o redistribut e an d repac k themselves . I f th e wal l movemen t awa y fro m th e
retained fill i s larg e enough, th e particles will be unable to maintain tension behin d th e
wall an d will tend to fall into any potential gap.
The soil-structur e interactio n mechanis m i s entirel y strain-driven . Previou s
analyses see m t o hav e opte d fo r calculations base d o n elasti c parameters , which wer e
chosen t o sui t th e operationa l range of stresse s an d strains (e.g . usin g a tangent shear
modulus, G t a n , whic h i s ver y different fro m the secant shear modulus G , Figur e 2c) .
s e c
However, th e past stres s history, th e recent strain path and strain increment following
the last reversal have a major influence o n the response.
Sands wil l follo w highl y non-linea r stress-strai n behaviou r (Figur e 2a), and ma y
show dilatanc y (strai n softening, Figur e 2a; volume increase , Figure 2b), an d further
exhibit significan t hysteresi s an d cyclic mobility (Figure s 2c and 2d) during these cyclic
deformations. A hyperbolic , hystereti c constitutiv e mode l i s probabl y th e mos t
suitable.
Preliminary dat a o n th e soi l respons e create d b y cyclin g betwee n stres s level s
(Figure 2c ) an d strai n o r displacemen t condition s (Figur e 2d ) hav e bee n presente d
respectively b y Basset t and Powrie (1992 ) an d Bassett et al. (1992) . Th e latter case i s
relevant t o thi s clas s o f proble m an d wil l b e referre d to i n greate r detai l durin g th e
current research programme.
The resulting stres s path followed by a n element o f soi l behind th e wall du e t o the
strain cycling i s highl y comple x - a combination o f compression , extensio n an d shear,
and will be investigated furthe r in due course. Th e element will be expected to undergo
some rotatio n i n principa l stresses , an d thi s wil l als o b e affecte d b y th e degre e o f
adhesion 6 obtaine d o n th e soil-wal l boundary . I n the field, thi s soil-wal l boundary
would be quit e rough , an d 8 m o D migh t b e close t o the critical state value fo r the soil,
c^'cnt. I n the event of a wedge failur e behind the wall, th e inclination o f the failure plane
will eithe r depen d o n th e valu e o f<!> ' i f th e conditio n o f maximu m stres s obliquit y
rules, or the lines of no extension, i n which cas e dilatancy \| / i s important.
The horizonta l eart h pressur e coefficien t i s likel y t o buil d up , mayb e eve n
approaching passive level s in really extreme cases. A s the abutment is cycle d int o and
away from the fill, thi s ratchetting ' behaviour will amplify th e bending moment s i n the
wall, an d significan t settlemen t i n th e regio n approachin g th e bridg e deck . Bot h o f
these mechanism s hav e implication s fo r th e serviceabilit y o f th e bridge . Crack s ma y
form in the abutment wall an d the road surface will settle.
Cohesionless soil s ar e most likel y t o b e use d i n practic e to construc t the approach
embankment, an d therefor e hav e bee n specifie d fo r thi s study . Ther e i s a greate r
probability o f damage to the abutment given the combination o f particular mobility an d
the stif f respons e o f compacted granular fills. Furthermore , th e peak angl e o f shearin g
resistance wil l b e highe r tha n typica l value s fo r clays , leadin g t o a potentially large r
lateral pressure coefficient an d more horizontal stress on the abutment wall.
Strain
Fig. 2a. Influence o f I o n x- y respons e Fig
d . 2b. Influence o f I o n volume chang e
d
• Strai n Strai n
Fig. 2c. Constan t stress cyclin g Fig. 2d. Constan t strain cycling
The influence o f number of cycles N o f deformation on the crucial parameters will

be of interest to the designer. D o th e wall bendin g moment s and pressures approach a
constant value ? I s th e settlemen t behin d th e wal l limite d afte r a specifi c numbe r of
cycles? I s there any variation in the behaviour if the abutment was initially perturbed in
an active (away from the retained fill) o r a passive (toward s the retained fill) sense ?
Table 2: Prototyp e specification an d scaling factors
Variable Field prototype Scale Centrifug e model
Bridge span 50-150 m 1/60 200 mm

Cyclic movement +/- 5 mm diurnally, 1/60 +/- 0.1 mm diurnally,
+/- 30 mm annually +/- 0.5 mm annually
(also test +/- 1 mm )
Retained height of fill 6m 1/60 100 mm
Embedment for piled wall 6m 1/60 100 mm
Piled wall flexural stiffnes s 1.5 10 kNm /m
5 2
1/60 3
0.69 kNm /m 2
Spread bas e wal l flexura l 1.5 WkNmVm 1/60 3

6.9 kNm /m
2
stiffness
Deck flexural stiffnes s 4.5 10 kNm /m
6 2
1/60 3
20.8 kNm /m 2
Sand particle size 5.4-9 mm 1/60 0.09-0.15 mm

Piled wall thicknes s 1/60 4.92 mm
Depth of deck 1/60 18.2 mm
Bending moment / m width 1/60 2
Shear force / m width 1/60

Lateral pressure 1
4 Structura l desig n o f the centrifuge model
The dimension s o f th e fiel d prototype , together with th e centrifug e mode l equivalen t

are given i n Table 2. Sinc e these tests are to be carried out at a nominal acceleration of
60 gravities , then ther e is a linear scaling facto r o f 6 0 on al l mode l dimensions . Th e
factors for flexural stiffness, bendin g momen t and shear force per metre length o f wal l
are also quoted in Table 2.
The genera l arrangement o f th e centrifuge mode l i s show n i n Figur e 3. Th e wal l
was made out of 6061/T6 grade Dural and the deck out of mild steel. Ther e was a rigid
moment-carrying connectio n a t th e wall-dec k joint . Th e displacemen t cycle s wer e
carried out usin g a steppe r motor driving a n eccentric rotating cam system throug h a
vertical pin joint. Thi s joint was also supported against vertical movement downwards
(but not upwards) at the end of the deck.
A ful l 360 ° rotatio n of the cam, without allowing fo r the small degree of backlash,
gives pea k t o pea k displacement s o f +/- 1 m m a t mode l scal e (1/60) , varyin g
sinusoidally, whic h become s +/ - 6 0 m m a t prototype scale . Th e cycle s appea r to b e

uniform an d repeatable a t each se t o f nomina l displacements . Fo r cam rotations less
than 360 ° , e.g . fo r +/ - 0. 1 m m (+/ - 6 m m prototype ) an d +/ - 0. 5 m m (+/ - 3 0 m m
prototype), the 'waveform' was sawtoot h in shape.
Concrete o r steel/concret e
composite dec k Concrete o r steel/concret e
composite dec k
Drained Pile ca p / rigid abutment-dec k 6m

cohesionless connection Pile ca p / rigi d abutment-dec k
Drained
in-situ soi l connection
cohesionless
or backfil l
backfill
material "Steel pil e abutmen t
material
- -
77r Reinforced concret e abutmen t
•
No momen t concret e connectio n
cohesionless soil foundatio n
cohesionless soi l foundatio n
M o d e l dimension s i n m m ( x 6 0 @ prototyp e scal e )
Fig. 3. Layou t of the centrifuge model
5 Modellin g th e soil
The choic e o f grai n siz e fo r th e mode l soi l i s critica l when th e mechanism s tha t will
obtain at particulate level are to be replicated. Th e smallest grai n size that it is practical
to wor k wit h i s silic a sand , 100/17 0 grad e (fractio n E) , with a particl e siz e rangin g
between 9 0 - 15 0 |im ( 5 . 4 - 9 m m at prototype scale). Therefore , for the initial se t of
strain cycles a t +/- 6 mm , extensiv e repackin g an d vertical settlement i s les s likel y t o
occur behind the wall.
The centrifug e model s wil l b e mad e fro m dr y san d sinc e th e strai n cyclin g i s s o
slow that pore pressure response is unlikely t o be significant - in contrast to the build up
associated wit h dynami c event s suc h a s earthquakes. I t is though t t o b e importan t to
isolate th e soil behaviour at particulate level a t this stage , given th e complications that
will arise with maintaining a n elevated wate r table behind th e wall. Shoul d significan t
leakage occur, the balance of the centrifuge is in jeopardy.
The mode l wa s prepare d by rainin g san d int o th e centrifug e strongbo x i n orde r to
achieve a relative densit y I o f abou t 80% . Th e I o f thi s fil l wil l hav e a major
d d
influence o n the soil-structure interaction. A zone of loose fill behind th e abutment will
be les s likel y t o caus e latera l pressure s t o buil d u p toward s passiv e levels . Th e
Americans favou r thi s method , an d us e a n approac h sla b t o spa n th e zon e tha t ha s
settled i n behind th e abutment. Othe r authorities prefer to specify a denser, compacted
fill.
The effec t o f differen t densitie s o f th e fil l behin d th e wal l wil l b e on e o f th e
variables examined in this series of tests. Relativ e density i s also an important indicator
of th e potentia l dilatanc y \\f of th e soil , whic h wil l elevat e th e mobilise d angl e o f
friction, tf mob wel l above the critical state value tf i (usuall y abou t 33° for quartz), s o
cr t
that in the extreme case (J)' b become s (J>' x.

mo ma
^'max M ' o i t + A v (1 )
A i s a constant whic h i s ofte n take n a s unity , althoug h Bolto n (1986 ) recommend s a

value of 0. 8 take n from back analysis o f serie s of triaxial and plane strai n test data. \\f
may b e determined, i n degrees , fro m the following formula e (Bolton , 1986 ) and it can
be see n tha t the dilatanc y i s increase d a t lo w mea n effectiv e stres s p ' (i n kPa ) and
when I i s high :
d
=\j/ 3 { I (10 - I n p') - 1 } (triaxial

d ) (2a )
=\|f 5 { I (10 - I n p') - 1 } (plan
d e strain) (2b )
A highe r peak value o f passive eart h pressure, with coefficient K p i s the n implied (at
the point when the major principal stress direction has rotated to the horizontal):
=
Kp ( 1 + sin / ( l - s i n 4»' mob ) (3 )
increasing a s tf approache s <|>' before droppin g of f t o ^

mob max (Figur e 2a), with the
CTlt
associated variation in lateral earth pressures and wall bendin g moments .
6 Tes t procedures
The wall wa s hel d in position whil e the san d was poure d at 1 g, an d s o th e stres s path
did no t entirely resembl e th e full scal e equivalent . Crai g (1984) note d tha t installatio n
of mode l pile s at the correct stress levels was far more critical for investigation o f axial
response tha n for lateral behaviour, which i s th e mode o f perturbation applied here . A
typical prototyp e horizonta l stres s profil e actin g o n th e abutmen t wal l followin g
compaction of the fill during construction is given in Figure 4.
Horizontal Stres s
Passive
c = v'2. p.y/%
Soil
Depth
Active
Fig. 4. Distributio n of horizontal stresses in compacted backfill i n the field
Following assembl y o f the actuator and connection of the displacement transducers

and bending moment strain gauge bridges, the model was loaded onto the centrifuge
and all systems were checked again from the centrifuge control room. Th e package was
then subjected to a brief load cycle created by accelerating the centrifuge model to 10 0
g s o that the soil was overstressed by a factor of 1.6 7 relativ e to the values expected at
a nomina l acceleratio n o f 6 0 g , an d th e eart h pressur e coefficien t wa s artificiall y
increased.
Three sets of displacement cycles were carried out:
(a) N = 5 0 cycles at +/- 6 mm, prototyp e scale

(b) N = 7 5 cycles at +/- 30 mm, prototyp e scale
(c) N = 10 0 cycles at +/- 60 mm, prototyp e scale (no t reported here).
In eac h case , a n initia l displacemen t o f th e sam e amoun t a s th e subsequen t

displacement cycles was carried out in either an active sense (awa y from the fill) or in a
passive sens e (toward s the fill). Dat a was take n from bending momen t transducer s i n
the wall s an d deck , togethe r wit h th e axia l loa d cel l i n th e dec k an d variou s
displacement transducers.
In thi s paper , a linea r approximatio n o f th e bendin g momen t diagra m wil l b e
integrated twice to give a profile of displacement down the wall, whic h may be fitted to
boundary condition s obtaine d fro m th e displacemen t transducer s t o chec k interna l
consistency o f th e data . Similarly , th e bending momen t profile ma y b e differentiate d
once to give an approximate shear force distribution for comparison with the axial load
reading a t dec k level . I n th e foreseeabl e future , th e dat a wil l b e fitte d b y eithe r a
polynomial o r a spline, whereupo n approximat e pressur e distributions will be obtained
by differentiating twice .
Centrifugal modelling of abutments 2 5 9
7 Preliminar y results
Preliminary analysi s o f dat a from two centrifug e mode l test s conducte d last week wil l
be presented below. Al l data is presented at prototype scales.
50 cycles: active initiation: +/-5m m excitation
Bending Momen t (kNm/m) Shea r Force (kN/m) Displacemen t (mm)

-25 0 5 0 10 0 -12 50 12 55 0 - 5
— I —I 1 1 1 1 1 1 I 1 1 1 1
1 I 1
I 1
12"
(a) Bendin g moment (b ) Shea r force (c ) Displacemen t
Key: - » - - N l -*_+N l -*--N2 6 _*_+N2 6 -4--N5 0 _^_+N5 0
N = no . of cycles
+ = displacemen t towards retained soil (passive)
- = displacemen t away from retained soil (active)
Sign convention: ^ — N
-ve •
Fig. 5 . Dat a from th e activ e initiation o f a smoot h wal l ove r 5 0 cycles o f +/ - 6 m m

7 5 c y c l e s : a c t i v e initiation : + / - 3 0 m m e x c i t a t i o n
Bending Momen t (kNrn/m) Shea r Force (kN/m) Displacemen t (mm)
6—
12
(a) Bendin g moment (b) Shea r force (c) Displacemen t
Key: -Nl .+N1 -N40 ) < +N4 0 -N75 , i . +N7 5
N = no . of cycles
- = displacemen t towards retained soil (passive)
+ = displacemen t away from retained soil (active)
Sign convention: ^
-ve •
Fig. 6. Dat a from the active initiation of a smooth wall over 75 cycles of +/- 30 mm
7.1 Smal l displacement behaviour

The bendin g momen t an d shear force per unit width , togethe r with wal l displacemen t
are give n i n Figure s 5a, 5b an d 5c, for 50 cycles o f +/ - 6 m m excitation at the deck.
There appeare d to b e minima l chang e i n th e bendin g moment s an d shear forces wit h
increasing numbers of displacement cycles, and also not much difference betwee n dat a
from bot h th e activ e an d the passiv e mode s o f deformation . Th e poin t a t which th e
bending moment s begi n t o amplify wit h increasin g numbers of cycle s require s further
investigation, bu t i t woul d see m tha t th e multipl e diurna l cycle s o f expansio n an d
contraction are unlikely t o cause significant distres s to the wall. Th e maximum vertical
settlement in the fill behind the wall was less than 12 mm for a zone 0.4 m long.
7.2 Larg e displacement behaviour

75 cycle s of +/- 30 mm displacement created a wedge formatio n in the sand behind the
abutment wall. Fo r the test in which excitation was initiated by an active displacement,
the wedg e forme d ove r almos t th e ful l heigh t o f th e retaine d fil l a t a n angl e o f
approximately 15 ° t o the vertical, and this wa s associate d with significan t settlemen t
of 210 mm at sub-pavement level .
The number of cycles di d not significantly affec t th e bending moment s (Figur e 6a )
generated by active displacemen t of the wall (indee d the peak moment occurring at the
1st cycl e wa s th e greatest) . However , th e pea k bendin g moment s du e t o successiv e
passive cycle s increase d quite noticeabl y an d migrated to a shallower depth , implyin g
that th e centroi d o f latera l pressure was als o movin g upwards . Th e cumulativ e an d
incremental escalatio n i n bendin g momen t togethe r wit h a n averag e % increase per
cycle ove r 7 5 cycle s i s summarise d i n Tabl e 3 . Afte r 4 0 cycles , th e rat e o f
accumulation is significantly reduced , but not eliminated altogether.
The shea r force diagra m (Figure 6b) demonstrate s the interna l consistency o f th e
data. A linear fit t o the bending momen t profile ha s been differentiated t o give a shear
force distribution for 1, 40 and 75 cycles of active and passive excitation. Thes e values
are compared with the axial load measured in the bridge deck by a series of transducers.
Agreement seems to be excellent.
Table 3. Increas e in bending moment for N cycles in the passive mode; active initiation
Cycle No. % increase in bending moment

(N) cumulativ e incremen t averag e % / cycle
2 9. 0 9. 0 9. 0
5 14.4 5.4 1.8
10 18.5 4.1 0.8
20 23.9 5.4 0.5
40 29.9 6.0 0.3
75 31.6 1.7 0.05
The displacement profil e o f the wall (Figur e 6c) confirms th e double integratio n of
the linear fit t o the bending moment s (value s measure d directly from the displacemen t
transducers ar e omitted fo r clarity) . I t is importan t to not e tha t there is neithe r rigid
body rotation , translation, shear force no r bending momen t a t the ti p o f th e abutment
wall.
An initial sca n of the data from the second centrifuge model tes t in which th e initial
excitation wa s i n th e passiv e directio n indicate s ver y simila r bendin g moment s i n
magnitude an d shape , an d comparabl e value s o f compressiv e loa d i n th e deck .
However, furthe r detaile d analysi s wil l b e require d t o explai n wh y th e wedg e
mechanism was smaller, both in depth and the volume of 'mobile' fill.
8 Futur e work
Aside fro m mor e detaile d analysi s o f th e dat a presente d here , th e variable s t o b e

investigated i n these and future centrifuge model tests from this series will be:
(a) densit y o f fill

(b) method s of compaction
(c) directio n of initial perturbation
(d) soil-wal l adhesio n
(e) flexura l stiffness o f the wall.
Amendments ar e planned t o the existing tes t package to giv e mor e direct readings
of pressure on the wall, and some information about in-situ, smal l strain soil stiffness .
Numerical analyses will be conducted in parallel to these tests to calibrate the model
data, with assistance from soil parameters obtained through plane strain stress path tests
that have replicated the stress history experienced by soil elements i n the centrifuge.
9 Conclusion s
The soil-structur e interaction for a n integral bridg e i s highl y comple x an d will no t be

well modelled b y elasticity. Severa l aspects must be considered to explain cohesionles s
soil behaviou r du e t o cycli c strains : particl e repacking, dilatancy , hysteresis , cycli c
mobility, an d stress paths due to repetitive compression, shear and extension.
Centrifuge mode l test s offe r a cost-effectiv e opportunit y o f investigatin g thes e
modes o f behaviou r by producin g dat a of axia l deck load , wal l bendin g moment s and
displacements, a s wel l a s exhibitin g th e deformatio n an d failur e mechanisms . Ke y
parameters may also be varied as required.
Large strai n cycle s creat e a n increas e i n wal l bendin g momen t combine d wit h a
clear wedge shape d failure mechanism behind th e wall, whic h i s associate d with sever e
slumping (0.2 1 m ) o f th e backfill . Afte r 7 5 cycle s a t +/ - 3 0 m m horizonta l
displacement a t the wall-deck joint, th e peak values o f bendin g momen t ha d increased
by nearly 32%, and the rate of accumulation had not completely levelled out. However ,
the change was minimal for the smaller cycles of +/- 6 mm which replicat e the extreme
diurnal variation.
10 Acknowledgement s
The firs t autho r of thi s repor t is employe d b y Cambridg e Universit y an d th e secon d

author is i n receip t of a grant from th e SERC . Th e wor k reported herein wa s carried
out under a contract placed on them by the Transport Research Laboratory. Th e views
expressed are not necessarily thos e of the Department of Transport.
The author s ar e also gratefu l t o inpu t fro m D r Dere k Carde r a t TRL , and t o th e
design and technical staf f at the Cambridge Geotechnical Centrifug e who helped t o turn
the ideas int o reality , an d to Dr Charles Ng fo r reviewing th e manuscrip t an d withou t
whom th e data would not have been ready for this paper.
References
Bassett, R.H. and Powrie, W. (1992 ) Th e behaviour of granular materials. Cours e for
industry. MAC E Centre, Imperia l College .
Bassett, R.H., Dalili, A. and England, G.L. (1992 ) Stres s changes in granular materials
subjected t o controlle d cycli c strains . Cours e fo r industry . MAC E Centre ,
Imperial College .
Bolton, M.D. (1991 ) Geotechnica l stres s analysis fo r bridge abutment design.
Transport & Road Research Laboratory Contractor Report, CR270.
Bolton, M.D. (1986 ) Th e strength and dilatancy of sands. Geotechniqu e 36 , No. 1 ,
65-78.
Bransby, M.F. (1992 ) Pile d foundations adjacen t to surcharge loads.
Cambridge University 1s t year Report. 50p .
Bransby, M.F. (1993 ) Centrifug e test investigation o f the buttonhole foundatio n
technique. Dat a Report. 33p .
Craig, W.H. (1984 ) Installatio n studie s fo r model piles, in The application o f
centrifuge modellin g to geotechnical design , (ed . W.H. Craig), Engineerin g
Department, Manchester, pp . 440-455.
Ellis, E.A. (1993 ) Latera l loading o f bridge abutment piles due to soil movement .
Hambly, E.C. (1991 ) Bridg e Deck Behaviour . 2n d edition, E & FN Spon., UK.
Norrish, A.R.M.. (1993 ) Soil-structur e interaction in rigid frame bridges.
Pritchard, B. (1992 ) Advantage s of bridge deck continuity, i n Bridge Design for
Economy and Durability. Thoma s Telford Publications, London, UK.
Schofield, A.N . (1980 ) Cambridg e University Geotechnica l Centrifug e Operations.
20th Rankine lecture. Geotechniqu e 30 , No. 3 , 227-268.
Springman, S.M. (1989 ) Latera l loading o n piles due to simulated embankmen t
construction. Cambridg e University Ph D thesis .
Springman, S.M. an d Bolton, M.D. (1990 ) Th e effect o f surcharge loading adjacen t to
piles. Transpor t & Road Research Laboratory Contractor Report, CR196.
Sun, H.W. (1989 ) Groun d deformation mechanism s fo r soil-structure interaction.
Cambridge University Ph D thesis.
23 TH E CONTRIBUTIO N O F
THE ROA D PAVEMEN T I N
A B S O R B I N G HORIZONTA L
MOVEMENTS I N A N
EMBANKMENT
A.M. LO W
Ove Am p & Partners, London, U K
Abstract
This pape r r e p o r t s t h e r e s u l t s o f a F i n i t e Elemen t compute r a n a l y s i s
o f a b r i d g e a b u t m e n t t o g e t h e r w i t h i t s a p p r o a c h embankmen t a n d t h e
n a t u r a l g r o u n d . Th e abutmen t i s s u b j e c t e d t o r e v e r s i n g h o r i z o n t a l
d i s p l a c e m e n t s . Tw o r u n s wer e made , on e w i t h an d on e w i t h o u t t h e
a d d i t i o n a l s t i f f n e s s o f t h e r o a d p a v e m e n t o n t h e e m b a n k m e n t . Th e
d i f f e r e n c e s i n t h e r e s u l t s o f t h e tw o r u n s a r e d i s c u s s e d .
Keywords: Abutment , Pavement , F i n i t e elemen t a n a l y s i s . Temperatur e
movements, Mohr-Coulom b
1 Introduction
B r i d g e d e s i g n e r s f a c e t h e sam e c h o i c e a s t h e d e s i g n e r s o f r a i l w a y
t r a c k s an d highwa y pavements : Shoul d th e effec t o f temperatur e
s t r a i n s b e c o n t r o l l e d w i t h movemen t j o i n t s a t c h o s e n c e n t r e s o r
should th e s t r a i n s b e r e s i s t e d b y e l a s t i c s t r e s s e s , r e s u l t i n g i n
l a r g e e n d r e a c t i o n s ? A s f a r a s I kno w t h o s e a u t h o r i t i e s w h i c h h a v e
b u i l t l o n g i n t e g r a l b r i d g e s ( s a y , l o n g e r t h a n 35m ) h a v e p r o v i d e d e n d
d e t a i l s w i t h a t r a n s i t i o n s l a b s p a n n i n g o v e r uncompacte d f i l l an d a
j o i n t somewher e betwee n t h e b r i d g e an d t h e r o a d pavement . T h i s
a n a l y t i c a l stud y consider s th e consequence s o f followin g th e othe r
route.
A s t u d y w h i c h wa s r e p o r t e d l a s t y e a r . Lo w ( 1 9 9 2 ) , i s r e w o r k e d a n d
e x t e n d e d b y c o m p a r i n g t h e b e h a v i o u r o f t h e abutmen t w i t h an d w i t h o u t
a r o a d pavemen t behin d i t .
2 The jointless bridge
We c a n s t a r t w i t h som e s i m p l e c a l c u l a t i o n s : Th e temperatur e o f a
c o n c r e t e b r i d g e migh t r i s e b y 27° C betwee n i t s c o n s t r u c t i o n i n t h e
w i n t e r an d i t s h o t t e s t da y i n s e r v i c e . I f i t s lengt h i s full y
r e s t r a i n e d an d i t s s e c t i o n i s unifor m t h e n t h e s t r e s s i n c o m p r e s s i o n
w o u l d r i s e a b o u t 1 0 MPa . T h i s i s h i g h s t r e s s t o a b s o r b o n t o p o f t h e
dead an d l i v e l o a d s t r e s s e s . I t w o u l d b e muc h e a s i e r t o c a t e r f o r
t h e f u l l y r e s t r a i n e d s t r e s s o f l e s s t h a n 5 MP a w h i c h r e s u l t s f r o m t h e
use o f lightweigh t c o n c r e t e .

266 Low
Pavement
Concrete Fill
li
8<
o\
o\
o/
os
o/
OV
o\
o/
OS
ox
OS
<iki<i<i<i<i<i<i^<i<i^<iq
'T- T I
IS 25 SO 63
X(m)
F i g . 1 . F i n i t e e l e m e n t mode l
F o r a s t e e l / c o n c r e t e composit e b r i d g e t h e t e m p e r a t u r e r i s e migh t
b e 3 5 ° C g i v i n g a f u l l y r e s t r a i n e d s t r e s s o f 8 6 MPa . T h i s a l s o i s
l a r g e i f th e s t e e l s t a r t s i n compressio n although , wit h c a r e , a
h a u n c h e d bea m s h a p e c a n b e c h o s e n w h i c h g r e a t l y r e d u c e s t h o s e
t e m p e r a t u r e c o m p r e s s i o n s whic h combin e wit h e x i s t i n g c o m p r e s s i o n s .
In p r a c t i c e t h e s e value s coul d onl y b e r e l e v a n t i f th e abutment s
were s e c u r e d t o r o c k f o u n d a t i o n s . Abutment s wit h s o i l foundation s
p r o v i d e p a r t i a l r e s t r a i n t t o t h e d e c k s , t h e y w i l l mov e u n d e r
t e m p e r a t u r e c h a n g e s a n d t h e s t r e s s e s w i l l b e muc h l o w e r . I t i s th e
i n t e r a c t i o n b e t w e e n t h e d e c k f o r c e s , abutmen t movement s an d t h e
s t r e s s e s an d s t r a i n s i n t h e s o i l t h a t I hav e bee n s t u d y i n g wit h a
f i n i t e elemen t a n a l y s i s .
3 Finit e element analysis
F i g . 1 show s a f i n i t e e l e m e n t h a l f mode l o f t h e s o i l a r o u n d a
s y m m e t r i c a l t w o - s p a n o v e r b r i d g e f o r a d u a l t h r e e l a n e m o t o r w a y . Th e
o v e r a l l l e n g t h o f t h e c o n c r e t e s t r u c t u r e i s 36 m a n d t h e r e i s 60 m o f
embankment b e h i n d t h e a b u t m e n t . T h e embankmen t i s 6 . 2 m h i g h o f w h i c h
t h e t o p 1 m i s d e s c r i b e d a s " p a v e m e n t " . Th e a b u t m e n t i s a c o n c r e t e
w a l l 1 m t h i c k an d 12.2 m h i g h , i n c l u d i n g i t s f o u n d a t i o n s . Bedroc k i s
26m b e l o w r o a d l e v e l .
L a t e r a l d i s p l a c e m e n t s a r e a p p l i e d 0.5 m belo w t h e t o p o f t h e
a b u t m e n t w a l l . Th e m o d e l i s a n a l y s e d w i t h SAF E ( S o i l A n a l y s i s u s i n g
F i n i t e E l e m e n t s ) whic h i s a n o n - l i n e a r progra m fro m Oasy s L t d w i t h
Horizontal movements in an embankment 267
2500
100
Displacement int o embankment m m
Run 1 Ru n2
Fig.2. Force/displacemen t plo t
s e v e r a l option s f o r d i f f e r e n t s o i l models . I t wa s r u n w i t h
e l a s t i c - p l a s t i c M o h r - C o u l o m b m o d e l s f o r t h e g r o u n d , t h e embankmen t
f i l l an d t h e pavement . I n i t i a l compactio n p r e s s u r e s wer e s p e c i f i e d .
The p r i n c i p a l m a t e r i a l p r o p e r t i e s assume d a r e :
Zone E ( k P a ) <t> ' c ' (kPa )

Concrete 2 3000000 0
Pavement 4 100000 0 3 8 20 0
Fill3 8000 0 3 60
N a t u r a l groun d 1 4000 03 40
Two r u n s w e r e m a d e . Th e o n l y d i f f e r e n c e b e t w e e n t h e t w o r u n s wa s
t h a t f o r Ru n 2 t h e m a t e r i a l p r o p e r t i e s f o r t h e p a v e m e n t z o n e w e r e
c h a n g e d t o t h o s e f o r t h e f i l l z o n e . Henc e t h e mode l r e p r e s e n t e d a n
earthwork wit h n o roa d o n top . F i g . 2 show s t h e f o r c e / d i s p l a c e m e n t
p l o t f o r e a c h r u n . T h e d i s p l a c e m e n t wa s i m p o s e d i n 6m m s t e p s , a
f o r w a r d movemen t o f f o u r s t e p s , awa y f r o m t h e e m b a n k m e n t , a n d t h e n a
b a c k w a r d movemen t o f e i g h t e e n s t e p s (108mm ) i n t o t h e e m b a n k m e n t . Th e
p l o t s o f e a c h o f t h e f o u r m o v e m e n t s i n d i c a t e som e d e g r e e o f p l a s t i c
softening a f t e r a n i n i t i a l e l a s t i c response . I n bot h th e forwar d
movements t h e l i n e a r , e l a s t i c r a n g e a p p e a r s t o f i n i s h somewher e n e a r
t h e en d o f t h e f i r s t s t e p an d t h e n t h e e f f e c t o f y i e l d i s q u i t e
marked. I n t h e b a c k w a r d s movement s t h e d e v i a t i o n fro m l i n e a r
behaviour i s ver y gradual . I n bot h run s s t r i c t l i n e a r i t y i s l o s t
a f t e r f o u r o r f i v e s t e p s b u t i n t h e n e x t tw o o r t h r e e s t e p s t h e
d e v i a t i o n i s s l i g h t an d p r o b a b l y r e p r e s e n t s a c o n d i t i o n whic h c a n b e
a c c e p t e d i n f r e q u e n t l y b y t h e embankmen t w i t h o u t i m p a i r i n g i t s
performance.
268 Low
_3S_ _3S_
7.5 3$_ 3 4 3| _ Yielde d
% %
5.0 _ % %
Y(m)
% %
% % %
%
T"
0_
-2.5 _ J
r 1^
17.5 20.0 25.0 27.5
F i g . 3 E x t e n t o f y i e l d i n g a t t h e en d o f Ru n 1
T h e p u r p o s e o f t h e s t u d y i s t o d e r i v e som e i n d i c a t i o n o f t h e
movement r a n g e w h i c h c a n b e s u s t a i n e d i n s e r v i c e s o i t i s i m p o r t a n t
t o u n d e r s t a n d t h e p h y s i c a l s i t u a t i o n whic h r e s u l t s i n d i f f e r e n t
d e g r e e s o f d e v i a t i o n fro m t h e e x t r a p o l a t i o n o f t h e l i n e a r b e h a v i o u r .
T h e SAF E p r o g r a m o f f e r s man y f o r m a t s f o r d i s p l a y i n g t h e v a r i a t i o n o f
s t r e s s e s a n d o t h e r v a r i a b l e s w i t h i n t h e s o i l m a s s . Th e numeri c
d i s p l a y o f "ph i m o b i l i s e d " g i v e s a r e c o r d o f t h e e x t e n t o f y i e l d i n g
a t an y s t a g e ( F i g . 3 ) . P h i - m o b i l i s e d i s th e g r e a t e s t f r i c t i o n angl e
( d e v i a t i o n o f t h e f o r c e r e s u l t a n t fro m t h e n o r m a l ) o n an y p l a n e
t h r o u g h a g i v e n p o i n t . Wher e yieldin g i s occurrin g i t s valu e i s th e
l i m i t i n g p h i o f t h e s o i l s o z o n e s o f y i e l d i n g c a n b e i d e n t i f i e d . Th e
e x t e n t o f t h e y i e l d e d z o n e s i s a n i n d i c a t i o n o f t h e d e g r e e o f damag e
being s u f f e r e d b y t h e groun d bu t i t i s a n i m p e r f e c t i n d i c a t i o n
b e c a u s e i t d o e s n o t sho w t h e i n t e n s i t y ( r a t e o f s t r a i n i n g ) o f t h e
y i e l d . Thi s i s onl y r e c o r d e d b y th e progra m a s th e r e s u l t i n g
cumulative displacemen t s o th e load/displacemen t plot s a r e probabl y
t h e b e s t a v a i l a b l e i n d i c a t o r o f damage .
Horizontal movements in an embankment 2 6 9
FORCE RANG E k N / m
IN DEC K A
F i g . 4 Force/displacemen t i n t e r a c t i o n diagra m
4 Th e results
The p r i n c i p l e r e s u l t c a n b e s e e n i n F i g . 2 . Th e exten t o f th e linea r

r e s p o n s e i s s i m i l a r f o r b o t h r u n s a n d i s a b o u t 36mm . I t h a d b e e n
c o n j e c t u r e d t h a t t h e s t i f f n e s s o f t h e pavemen t woul d p r o v i d e a f o r c e
p a t h t o d e l i v e r som e o f t h e a b u t m e n t t h r u s t i n t o t h e embankmen t w e l l
b a c k fro m t h e b r i d g e . F o r a g i v e n s t r e s s l e v e l t h e s t r a i n s woul d b e
mobilised ove r a longe r lengt h resultin g i n l a r g e r displacements .
The r e s u l t s d o n o t s u p p o r t t h i s c o n j e c t u r e , p r e s u m a b l y b e c a u s e t h e
a d d i t i o n a l s t i f f n e s s i n c r e a s e s t h e t o t a l f o r c e an d c o u n t e r s t h e
effect.
However t h e a d d i t i o n a l f o r c e i s u s e f u l i n t h e q u e s t f o r l o n g ,
j o i n t - f r e e b r i d g e s b e c a u s e i t double s t h e s t i f f n e s s o f t h e abutmen t
a n d t h u s i n c r e a s e s t h e p r o p o r t i o n o f t h e t e m p e r a t u r e movemen t t h a t
can b e a b s o r b e d i n dec k s t r a i n s .
The a b u t m e n t s t i f f n e s s e s f r o m t h e l i n e a r r a n g e o f t h e t w o r u n s a r e
4 3 . 1 kN/m m a n d 2 0 . 6 kN/mm . A c o n c r e t e b r i d g e d e c k m i g h t h a v e a mea n
c o n c r e t e s e c t i o n a l a r e a o f a b o u t 0 . 4 5 m /m an d b e s u b j e c t e d t o a
2
t e m p e r a t u r e r a n g e o f 4 7 ° C Hence , w i t h t h e p a v e m e n t , t h e maximu m
d e c k l e n g t h w h i c h c o u l d o p e r a t e w i t h i n t h e 36m m movemen t r a n g e w o u l d
b e 163m . W i t h o u t t h e pavemen t i t woul d b e 142m . Th e e q u i v a l e n t
f i g u r e s f o r a d e c k i n l i g h t w e i g h t c o n c r e t e w o u l d b e 313 m a n d 2 2 5 m .
270 Low
T h e d i a g r a m u s e d t o d e r i v e t h e s e v a l u e s i s show n i n F i g . 4 . I t
d e m o n s t r a t e s t h e i n t e r a c t i o n betwee n t h e f o r c e i n t h e d e c k an d t h e
displacement a t eac h abutment .
F i g . 5 H o r i z o n t a l d i s p l a c e m e n t s i n t h e pavemen t
a t t h e e n d o f Ru n 1
R e s u l t s wer e a l s o p l o t t e d f o r t h e v a r i a t i o n o f t h e h o r i z o n t a l
d i s p l a c e m e n t s b e h i n d t h e a b u t m e n t . F i g . 5 s h o w s t h e p l o t a t maximu m
d i s p l a c e m e n t f o r Ru n 1 . T h e e q u i v a l e n t p l o t f o r Ru n 2 i s s i m i l a r .
T h i s show s t h a t t h e s t r a i n ( t h e s l o p e o f t h e p l o t ) h a s n o t f a l l e n t o
z e r o a t t h e en d o f t h e mode l s o t h e mode l s i z e woul d a f f e c t t h e
stiffness. I n r e a l i t y t h e s t r e s s e s woul d d i s s i p a t e i n t o t h e t h i r d
d i m e n s i o n an d t h e n o n - l i n e a r b e h a v i o u r woul d a l s o r e d u c e t h e e f f e c t
o f mor e d i s t a n t s t r a i n s . Henc e a b i g g e r mode l woul d p r o b a b l y n o t
givea b e t t e r stiffness .
Horizontal movements in an embankment 27 1
5 Commentary
These r e s u l t s shoul d no t b e applie d t o bridg e design s withou t takin g

full accoun t o f th e effect s o f th e larg e r e s t r a i n t forces . I t shoul d
be r e c o g n i s e d t h a t t h e r e s t r a i n t f o r c e w i l l follo w a s t r a i g h t l i n e i n
p l a n b e t w e e n t h e e l a s t i c c e n t r e s o f t h e tw o a b u t m e n t s a n d c a r e i s
n e e d e d i n p l a n n i n g a ske w b r i d g e o r o n e w i t h a c u r v e d a l i g n m e n t . Th e
d i g g i n g o f s e r v i c e t r e n c h e s a c r o s s t h e highway , c l o s e behin d t h e
a b u t m e n t , w o u l d r e l e a s e som e o f t h e t e m p e r a t u r e r e s t r a i n t . Som e
c o n t r o l o v e r s u c h a c t i v i t i e s woul d b e r e q u i r e d .
6 Conclusions
The r e s u l t s o f t h i s s t u d y i n d i c a t e t h a t t h e i n c l u s i o n o f t h e
h o r i z o n t a l s t i f f n e s s o f t h e highwa y pavemen t i n t h e a n a l y s i s a l l o w s
j o i n t - f r e e b r i d g e s t o b e d e s i g n e d w h i c h a r e 15 % l o n g e r t h a n t h e y
would o t h e r w i s e b e . I n genera l lightweigh t concret e j o i n t - f r e e
b r i d g e d e c k s c a n b e l o n g e r an d i n p a r t i c u l a r t h e y d e r i v e a
p r o p o r t i o n a t e l y g r e a t e r b e n e f i t fro m t h e a d d i t i o n a l r e s t r a i n t du e t o
t h e r o a d p a v e m e n t . Th e i n c r e a s e i s a b o u t 40% .
7 Reference
Low, A.M . ( 1 9 9 2 ) C o n c e p t s i n t h e d e s i g n o f t h e abutmen t i n i n t e g r a l

bridges, i n B r i d g e D e s i g n f o r D u r a b i l i t y , E x p a n s i o n J o i n t s an d
C o n t i n u i t y (e d Stephe n R a g g e t t ) , T r a n s p o r t R e s e a r c h L a b o r a t o r y ,
Crowthorne.
24 TH E INTERACTIO N
B E T W E E N BRIDGE S AN D
EMBANKMENTS AN D TH E
USE O F RUN-O N SLAB S
S. THORBUR N
Thorburn Ltd, Glasgow, UK
Synopsis
The interactio n betwee n earthwor k embankment s an d bridge s ha s bee n o f interes t
to engineer s ove r man y year s becaus e o f th e incongruit y o f bridg e structure s
on pile d foundation s an d approac h embankment s restin g o n sof t cla y soils . Th e
incompatibility o f vertica l displacement s betwee n bridg e structure s and adjacent
embankments create s problem s i n respec t o f roa d pavemen t performanc e an d
maintenance.
Opportunity wa s take n durin g motorwa y constructio n i n Scotlan d i n 197 6 to
conduct extensiv e geotechnica l investigation s an d t o instal l instrumentatio n
to monito r th e performance s o f embankment s i n clos e proximit y t o bridg e
structures.
The pape r examine s th e conclusion s whic h ma y b e draw n i n respec t o f
incompatibility o f movements an d the appropriatenes s of run-on slabs. Th e effects o f
lateral soi l displacement s o n a bridge structure also ar e presented in the paper.
Keywords : Interaction , Embankments , Bridges , Clays , Geotechnical ,
Instrumentation, Run-on-slabs , Displacements .
1 Introductio n
The tw o prim e structure s i n th e desig n o f motorway s an d trun k road s are

embankments an d bridges.
These component s ar e integral wit h eac h othe r i n respec t o f performanc e an d i t
is insufficient t o design them as separate elements.
Concrete o r stee l bridge s (hardware ) an d soi l embankment s (software ) interac t
with eac h othe r an d bridg e designer s canno t affor d t o ignor e th e interactiv e
effects.
The subjec t o f interactio n i s covere d i n th e publicatio n o n th e Rea l Behaviou r
of Structure s publishe d b y th e Institutio n o f Structura l Engineer s i n Marc h
1989.
It i s th e experienc e o f th e autho r tha t ther e i s a n over-emphasi s o n th e
performances o f embankment s wit h lesse r attentio n bein g give n t o th e
performances of bridges.
The major problem confronting th e designer o f embankments t o be formed on very
soft cohesiv e soil s i s no t th e mechanic s o f stabilit y analyses,sinc e thes e ar e
well established , bu t rathe r th e difficult y o f estimatin g fo r th e purpose s o f
analysis th e probabl e magnitude s an d rate s o f dissipatio n o f th e exces s
porewater pressure s generate d b y th e constructio n o f th e embankments . Th e
magnitudes an d rate s o f dissipatio n mus t b e predicte d wit h reasonabl e accurac y

274 Thorburn
to permi t prope r definitio n i n contrac t document s o f th e sequenc e an d rate s o f

embankment construction . Contractor s canno t efficientl y pla n thei r majo r
earth-moving operation s i f programme s o f constructio n canno t b e formulate d o n
the basis of reliable geotechnical predictions.
Pause clause s shoul d b e incorporate d i n tende r documents an d the word pause
refers t o th e perio d o f tim e betwee n eac h stag e o f constructio n o f a n
embankment durin g whic h earth-movin g operation s ar e discontinue d t o permi t
excess porewate r pressure s t o dissipat e t o acceptabl e residua l values . Th e
extent t o whic h judgemen t i s use d i n paus e claus e prediction s depend s o n th e
availability o f characteristi c data , and , i f th e volum e o f prope r dat a i s
relatively smal l an d variabl e the n statistica l theor y canno t b e applie d an d
greater reliance must be placed on judgement.
A cas e histor y i s presente d i n thi s pape r for a high embankment founde d o n
post-glacial deposits whic h comprised very soft cars e clays.
Although th e expenditur e o n th e sit e investigatio n wor k fo r th e motorway
project wa s significant , th e variation s i n th e permeabilit y an d
compressibility characteristic s o f th e sof t alluviu m necessitate d considerabl e
judgement i n th e selectio n o f desig n parameters . Fiel d permeabilit y test s
were carrie d ou t t o determin e th e coefficient s o f consolidatio n t o b e use d fo r
the paus e claus e prediction s an d extensiv e instrumentatio n wa s installe d t o
permit the performances of the embankments and bridges to be monitored.
This paper examines th e performanc e o f an embankment and an adjacent bridge
structure constructe d som e 1 6 year s ag o an d discusse s th e relevanc e o f th e
case history to the design of run-on slabs.
2 Geolog y o f site
The geomorphi c histor y o f th e sit e i s derive d fro m glaciation , change s i n se a

level an d estuarin e activity . Th e superficia l deposit s ma y b e assigne d t o th e
following broad categories depending o n their mode of origin:
Post-glacial deposits (cars e clays)
Late-glacial deposits (laminate d silty clays)
Glacial deposit s
The post-glacia l deposit s ar e considere d t o b e o f estuarin e origi n an d hav e
limiting elevation s abov e ordnanc e datu m o f abou t 1 4 an d 4 2 metre s
respectively.
Fig 1 show s th e locatio n o f a n old shorelin e whic h form s a distinct physica l
land feature . Th e abrup t chang e i n groun d leve l define s th e existenc e o f
deposits o f late-glacial age and the post-glacial carse .
The cars e consist s o f relativel y featureles s ver y silt y clay s whic h ar e
Fig. 1 . Locatio n of old shoreline .

Interaction between bridges and embankments 27 5
analogous t o th e tida l mud-flat s currentl y expose d a t lo w tid e a t th e Fort h

River estuary . Th e cars e i s considere d t o b e lightl y over-consolidate d excep t
in th e surfac e layer s whic h generall y hav e a relativel y firme r consistenc y du e
to desiccation.
An organi c deposit know n a s sub-carse peat defines a n important emergence of
the land between th e late-glacia l an d th e post-glacia l submergence.
A comple x assemblag e o f marin e deposit s underli e th e sub-cars e pea t an d
consists essentially o f various gradations of silts and sands.
The late-glacia l deposit s whic h overl y th e glacia l til l consis t o f laminate d
silty clays , an d ar e typifie d b y alternatin g band s o f silt y cla y an d sil t an d
fine sand . The y ar e generall y foun d withi n depression s i n th e irregula r and
undulating surfac e of the glacial till and are over-consolidated.
3 Groun d condition s
An importan t requiremen t i n th e desig n o f earthwork s fo r motorway s i s a n
adequate definitio n o f th e stres s historie s an d undraine d strength s o f th e
superficial deposits , particularl y wher e hig h embankment s wil l b e constructe d
over sof t cohesiv e soils . Th e determinatio n o f thes e propertie s permit s a
rapid assessmen t o f th e probabilit y o f instabilit y o f embankment s an d focuse s
attention o n th e portion s o f lan d whic h requir e specia l consideratio n an d
comprehensive investigator y work . Th e constructio n measure s fo r avoidanc e o f
embankment failur e ca n caus e significan t delay s t o th e constructio n programme
and al l location s o f potentia l instabilit y mus t b e define d a t a n earl y stag e
in the design of the earthworks.
The undraine d shea r strength s o f th e post-glacia l an d late-glacia l deposit s
were obtaine d fro m undraine d triaxia l compressio n test s an d lowe r boun d
strengths were deduced from cone penetration tests (CPT), a s shown in Fig 2.
Where hig h embankments ar e to b e constructe d on sof t cla y soils , knowledg e i s
required o f th e in-sit u soi l permeabilitie s t o permi t calculatio n o f th e rate s
of dissipatio n o f th e exces s porewate r pressure s generate d b y th e embankmen t
loading. Thi s importan t informatio n fo r th e predictio n o f th e variation s i n
the facto r o f safet y agains t embankmen t failur e wit h tim e an d o n th e rate s o f
settlement of the embankments was obtained from piezometer measurements.
A critica l location fo r embankmen t stabilit y o n th e cars e wa s selecte d fro m a
study o f th e sit e investigatio n informatio n an d thre e row s o f piezometer s wer e
installed a t position s correspondin g t o th e centr e line , mi d slop e an d outwit h
the toe of the proposed embankment as shown in Fig 3.
Typical value s o f th e horizonta l an d vertica l permeabilitie s o f th e cla y soil s
are given in Fig 4.
The strengt h parameter s o f th e soil s i n term s o f effectiv e stresse s wer e
determined fro m consolidate d undraine d triaxia l compressio n test s wit h
porewater pressur e measurement s an d thi s informatio n wa s use d t o predic t th e
factors of safety agains t embankment failure.
The compressibilitie s o f th e sof t clay s wer e determine d fro m laborator y
consolidation test s o n specimen s prepare d from sample s obtaine d from
stationery thi n walle d pisto n samplers . Longter m laborator y consolidatio n
tests wer e als o carrie d ou t t o establis h th e relationship s require d fo r th e
determination of the secondary consolidation displacements .
276 Thorburn
PROBABLE LOWE R BOUND S O F UNDRAINE D SHEA R STRENGTH S (Su ) O F

COHESIVE DEPOSIT S DEDUCE D FRO M CON E PENETRATIO N TEST S
01 02 03 04 05 06 07 0
UNDRAINED SHEA R STRENGT H (kN/m )2
Fig. 2 . Undraine d shear strengths, deduce d from cone penetration tests.

Interaction between bridges and embankments 2 7 7
NORTH BOUN O CARRtAGEW / Y
PULVERISED FUE L AS H
ORAINAGE LAYE R
EXISTING GROUN D LEVE L
E5Z
1
PEA T BASAL LAYER!SAN D
SILTY CLA Y ' C'
~SANDY SIL T
Fig. 3 . Location of piezometers.
M876 BRIDG E 1 3
PLOT O F PERMEABILIT Y v DEPT H
PERMEABILITY ( m / s E C . )
-* - 5 - « - 7 - » - S -1 0
10 1 01 01 01 01 01 0
SILTY CLA Y A
SILTY CLAY'B '
7
L
5
HORIZONTAL. SILTY CLA Y "C "
2A, , 1
1
HORIZONTAL VERTICA L
SANDY SIL T
Fig. 4. Typical values of horizontal and vertical permeabilities.

278 Thorburn
4 Prediction s o f rat e o f settlement an d stabilit y o f embankment s

The Transpor t and Roa d Researc h Laboratory publishe d a report, Murra y (1974 )
which describe d a metho d o f analysin g th e magnitud e an d rat e o f settlemen t o n
multi-layers soil s resultin g fro m a proces s o f two-dimensiona l consolidatio n
together wit h porewate r pressur e generatio n an d dissipation . I n additio n t o
the facilit y fo r analysin g stratifie d soil s th e metho d coul d tak e accoun t o f
the variatio n o f soi l parameter s durin g consolidatio n togethe r wit h non-linea r
relationships.
The followin g integrant s o f settlemen t requir e consideratio n i n th e assessmen t
of long-term performance of embankments on soft clay soils .
a. Elasti c displacement s unde r embankment loadin g
b. Primar y consolidatio n displacement s du e t o increase s i n majo r principal
stresses wit h th e associate d generatio n an d dissipation o f exces s porewater
pressures.
c. Shea r displacement s cause d b y hig h ratio s o f applie d stres s t o limitin g
stress.
d. Secondar y consolidatio n displacements whic h commenc e durin g th e primary
phase and continue after dissipation o f all excess porewater pressures.
Sufficient laborator y and field tes t dat a were accumulate d to permit the displacements
described in a, band d t o be predicted together with th e probable porewater pressures
induced in the soils by the embankment loading.
As a n important aid to judgement meticulous examinatio n was made of continuou s
air-dried spli t sample s an d th e component s o f th e physica l structure s o f th e
clay soil s wer e carefull y identifie d an d recorded . Subsequen t examinatio n o f
colour photograph s o f th e continuou s sample s take n fro m th e firs t borehol e o f
each ro w o f piezomete r borehole s togethe r wit h th e record s o f th e natur e an d
thicknesses o f th e physica l feature s o f th e sample s permitte d a n assessmen t o f
the parameters to be used in the analyses.
The physica l mode l adopte d fo r th e analytica l wor k i s show n i n Fi g 5 together
with the selected soil characteristics.
The predicte d magnitud e an d rat e o f settlemen t unde r th e centrelin e o f th e 9
metres hig h pulverise d fue l as h (PFA ) embankment founde d o n th e sof t cars e i s
Fig. 5. Physica l model an d soil characteristics for analysis.

shown o n Fi g 6 , wit h an d withou t recognitio n o f th e influenc e o f secondar y

consolidation whic h wa s foun d t o b e onl y 6 % of th e maximu m tota l settlement .
It wa s assume d fo r th e purpose s o f th e predictiv e analyse s tha t th e contracto r
would wis h t o plac e th e fil l materia l expeditiousl y an d a n averag e rat e o f
construction o f 1.1 2 metre s pe r mont h wa s envisaged . I n realit y th e rat e o f
deposition o f th e PF A wa s variabl e an d average d onl y abou t 0.6 8 metre s per
month requirin g a 1.1 0 yea r perio d o f constructio n instea d o f 0.6 7 yea r a s
permitted by the contract documents.
The allowabl e rat e o f constructio n o f th e embankment s wa s base d o n th e
porewater pressure predictions calculate d b y th e TRR L metho d an d the minimu m
factor o f safet y agains t slop e instabilit y a t an y stag e o f constructio n i n
terms o f effectiv e stresses , assesse d from th e Bisho p (1955 ) metho d o f
analysis, wa s 1.50 . Th e rati o o f applie d shea r stres s assesse d from
Jurgenson (1934 ) t o th e limitin g stres s give n b y th e undraine d strengt h o f th e
carse locall y approache d unit y withou t an y evidenc e o f larg e shea r strains .
Some evidenc e o f shea r displacemen t a t suc h hig h stres s ratio s woul d hav e bee n
expected unles s th e stresse d zone s beneat h th e embankmen t generall y ha d
adequate reserve s o f strengt h an d wer e capabl e o f accommodatin g a loca l dee p
zone of plasticity.
5 Performance s of embankment s
The predicte d an d measure d rate s o f settlemen t o f th e 9 metre s hig h

embankments founde d o n th e cars e ar e show n i n Fi g 7 . Th e vertica l
displacements wer e measure d bot h b y precis e leve l an d settlemen t gauge s
comprising horizonta l tube s beneat h th e embankment . Th e origina l settlemen t
predictions ar e give n togethe r wit h th e prediction s amende d t o confor m wit h
the slower rate of construction achieved during the contract.
WITH SECONDAR Y CONSOLIDATION

• 00
MAX ! 7 5 2 mm AT 1 0 YEARS
MAX . 7 0 6 m m AT 1 0 YEARS
WITHOUT SECONDAR Y CONSOLIDATION
HEIGHT O F EMBANKMEN T
HEIGHT O F
EMBANKMENT ( m )
10
6.0
1.0
IT 1 — T — 30 AO
0
SO
0 1. 0 2. 0 TIME ( Y E A R S )
Fig. 6. Predicte d settlement under centreline of embankment.

280 Thorburn
8OO-1
DATE
Fig. 7. Predicte d and measured settlement
Fig 8 show s a typical recor d of measuremen t o f th e exces s porewate r pressures

generated relativ e t o th e actua l rat e o f depositio n o f th e fil l materia l ove r
a perio d slighd y i n exces s o f on e year . Thi s recor d fo r a piezomete r whic h
was installed at a dept h of 3 metres under the centreline of the embankment, shows the
rapid increase s in porewater pressures on application of embankment loading an d the
gradual dissipation wit h time.
The piezomete r measurement s indicat e tha t onl y smal l exces s porewate r
pressures wer e generate d b y th e constructio n o f th e initia l 2.5 0 metre s thic k
layer, an d wer e equivalen t t o abou t 0.3 0 metre s o f water . Th e loadin g
intensity impose d o n th e formatio n a t th e centrelin e o f th e embankmen t b y 2.5 0
metres o f pulverise d fue l as h i s approximatel y 3 5 kN/m . Th e recorde d maximum
2
excess porewate r pressur e o f 3 kN/m i s a relativel y smal l proportio n o f th e

2
applied stres s an d thi s behaviou r ha s bee n observe d wher e th e exces s porewate r

pressures wer e foun d t o b e smal l unti l th e embankmen t loadin g exceede d th e
general over-consolidation stres s on the soft cars e of about 35 kN/m .
2
6 Performance s of bridge s
The previou s section s havin g define d th e probabl e performance s o f typica l

embankments o n sof t cla y soil s se t th e backgroun d agains t whic h w e ca n examin e
the performances of bridge structures.
We learn much fro m th e performance s o f earthwor k embankment s in terms of
deformations an d i t ma y b e o f valu e t o bridg e designer s t o conside r th e
criteria liste d belo w whic h wer e derive d fro m th e fiel d measurement s describe d
in this paper.
1. Th e tota l settlemen t o f 9 metre s hig h embankment s o n sof t cla y soil s may
be expected to be at least 700mm.
MEASURED POR E WATE R PRESSUR E
ACTUAL CONSTRUCTIO N RAT E
ENVISAGED CONSTRUCTIO N RAT E
PRCOICTEO POR E WATE R PRESSUR E

ADJUSTED T O CONFOR M WIT H ACTUA L
CONSTRUCTION RAT E
Fig. 8 . Typica l excess porewater pressures.
2. Th e tim e take n to achiev e th e fina l settlemen t o f 9 metres high embankment s

on sof t clay soil s may be expected to be at least 2 years after completion o f the
embankment.
3. Latera l displacement s o f sof t cla y soil s unde r 9 metre s hig h embankment s
may be expected to be at least 7 0 to 150mm.
In order to gain an indication o f the situation immediatel y behin d a bridge abutment
where an embankmen t has been formed on soft clays, le t us consider th e semi-infinit e
half-plane o f homogeneou s cla y loade d asymmetrically . Le t us assum e tha t surfaces
of rupture develop in the soft cla y soi l and the mass rotate s clockwise o n a cylindrical
slip surface which i s o f uni t lengt h normal to the section shown in Fig 9. Al l norma l
reactions o n th e cylindrica l sli p surfac e pas s throug h the centre of rotatio n and all
tangential reactions are of magnitude (Su ) per unit length o f arc .
The mathematical solution for the worst case is
(q-p)= 5.53S u
Where Su is the undrained shear strengt h of the clay. (Schofiel d an d Wroth, 1968) .
With referenc e t o Fi g 2 , i f w e assum e tha t th e mea n undraine d shea r strengt h
of th e sof t clay s ar e 2 0 kN/m a n approximatio n t o th e maximu m heigh t o f a
2
PFA embankment befor e failur e occurs is

5.53 x 2 0 = 8 metres
14
Embankments ar e constructed , therefore , wit h 2 t o 1 longitudina l sid e slope s
in orde r t o ensur e factor s o f safet y rangin g fro m abou t 1. 3 t o 1. 5 an d th e
soil displacements whic h are experienced ar e shown in Fi g 10.
If embankmen t fil l materia l i s deposite d agains t a bridg e abutment ,
as show n i n Fi g 11(a) , th e sof t cla y soil s upo n whic h th e hig h embankmen t i s
constructed wil l approac h limi t stres s an d plasti c behaviou r o f th e sof t clay s
282 Thorburn
R s i na
<7
2Rotk
(SCHOFIELD AN D WROTH , 1968 )
Fig. 9. Assume d mode l o f slip surface.
will b e experienced . Th e lateral displacement s o f th e sof t clay s i n a plasti c

state impose latera l loads on the piles supporting the bridge abutments.
The latera l load s ar e considerabl e an d displacement s o f piled abutment s ar e
experienced, as shown in Figs 12 an d 13.
Advanced earthworks which involv e construction of embankments some one to two
years befor e constructio n of bridges ma y improv e embankmen t performance , bu t
Inclnomeler N * 2
.Inclinometer N * 3 Inclinomete r N *'
TT
HORIZONTAL DISPLACEMEN T AGAINST TIME
Fig. 10 . Inclinomete r curves for M876.

BRIDGE
DECK »
WEDGE O F
INFILL
EMBANKMENT FIL L
BEARING STRAT A
(a)
INTEGRAL
BRIDGE DEC K T RUN - ON SLAB T ROAD PAVEMEN T ,
BEARING STRAT A
Fig. 11 . Alternativ e methods of constructing embankments.
SOUTH ABUTHEN T NORTH ABUTHEN T

November 7 8 August 7$
10 20m m
TYPICAL BOREHOL E I
Fig. 12 . Displacemen t o f piled abutments of M876 Bridge 20, measured with

inclinometers.
284 Thorburn
Fig. 13 . Directio n of abutment movement, M876 Bridge 20, as indicated by

inclinometers.
do no t reliev e th e near-limi t stresse s induce d b y th e fina l infillin g o f th e

wedges behin d the bridge abutments.
The advance d earthwork s coul d b e extende d ove r individua l bridg e site s t o
pre-load th e sof t clay s an d reduc e th e effec t o f infillin g o f th e wedge s
behind th e abutments . Ther e would , however , b e a cos t an d tim e penalt y i n
adopting this solution, includin g th e problem of early land acquisition.
The alternativ e solutio n o f constructin g additiona l bridg e deck s supporte d o n
bank seat s an d th e constructio n o f latera l sid e slope s t o minimis e soi l
displacements, a s shown o n Fig 11(b) , ma y be adopted together with the use of 'soft '
piles to support bank seats.
The concept of run-on slabs was developed a s an adjunct to integral bridges.
Integral bridge s ma y b e describe d a s thos e constructe d without movemen t joint s
over th e entir e lengt h o f th e bridg e deck . I n orde r t o creat e a transitio n
between th e continuou s concret e bridg e dec k an d th e roa d pavements a t eithe r
end of the bridge, run-on slabs may be provided, as shown on Fig 11(b) .
Joints ar e require d betwee n run-o n slab s an d reinforce d concret e roa d
pavements i n orde r t o accommodat e movements , bu t th e interface s betwee n
bituminous roa d pavements an d run-o n slab s ar e no t provide d wit h specia l
joints.
Degradation o f th e roa d pavemen t a t thes e interface s i s inevitable , bu t
maintenance costs apparently are considered acceptable.
In situations where embankments hav e been formed on soft clays, Figll(a) , run-on
slabs canno t b e use d ove r th e length s o f wedge s o f infil l behin d bridg e
abutments sinc e the y d o no t eliminat e th e proble m o f soi l displacement s a t
depth.
In orde r to overcom e the problem o f soi l displacement s behin d bridg e abutments
the autho r propose d i n 197 6 th e solutio n o f usin g geotextile s an d individua l
piles to support the wedges o f fill.
It wa s propose d tha t the portions o f embankment s immediatel y behin d bridg e

abutments b e supporte d o n individual piles arranged such that th e percentage of total
embankment load carried by the piles would reduce with increasin g distance from th e
abutment. Th e bridges would be designed with norma l abutments without additiona l
side span s and the need for embankment pre-loading would be removed. A significan t
benefit wa s achieved by the removal of any restrictions on rate of infilling behind bridge
abutments.
A comprehensiv e descriptio n o f th e suppor t system wa s presente d by Rei d and
Buchanan (1984).
The Autho r does no t recommen d th e us e o f run-o n slab s i n situation s wher e
infill would be place d on soft clays immediatel y behin d bridg e abutments and where
large soi l displacement s woul d b e expected . Th e researc h being conducte d b y th e
University of Cambridge should be supported by Highway Authorities.
7 Reference s
Bishop, A.W . (1955) . Th e us e o f th e Sli p Circl e i n th e stabilit y analysi s o f

slopes. Geotechniqu e 5 (7), 17 .
Francis, E.H. , Forsyth , I.H. , Read, W.A. an d Armstrong, M. (1970 ) - Geolog y o f
Stirling District .
Hambly E.C. (1991 ) Bridg e Deck Behaviour . E&FNSpon .
Hvorslev, M.J . (1951 ) Tim e la g an d soi l permeabilit y i n groundwate r
observations. Bulleti n N o 36 , Waterway s Experimen t Station , Corps , o f
Engineering, Vicksburg, Mississippi .
Jurgenson, L . (1934) . Th e applicatio n o f theorie s o f elasticit y an d
plasticity t o foundatio n problems . Journa l o f th e Bosto n Societ y o f Civi l
Engineers.
Murray, R.T . (1973 ) Embankment s constructe d o n sof t foundations : Settlemen t
studies near Oxford. TRR L Repor t No 538.
Murray, R.T . (1974 ) Tw o dimensiona l analysi s o f settlemen t b y compute r
programme. TRR L Repor t No 6117.
Reid W.M . an d Buchana n N.W . (1984 ) Bridg e approac h suppor t piling .
Proceedings of conference o n advance s i n piling and ground treatment. Th e
Institution of Civil Engineers, March 1983.
Schofield, A . and Wroth, P. Critica l state soil mechanics. McGraw-Hill .
Sissons, J.B . (1971 ) Geomorpholog y an d foundatio n condition s aroun d
Grangemouth. Quarterl y Journal, Engineerin g Geology 3,183-191 .
Thorburn, S . (1976 ) Th e stati c penetratio n tes t an d th e ultimat e resistanc e
of drive n piles i n fine-graine d non-cohesiv e soils . Th e Structura l Engineer
54, 6.
Thorburn, S . (1988 ) Soi l Interation , Th e Rea l Behaviou r o f Structures.
Institutions o f Structural an d Civil Engineers .
R E P O R T O N SESSIO N EIGH T
D.W. QUINIO N
Dr Springma n describe d th e inference s fro m th e recen t centrifug e testin g

examination o f th e soil s behaviour behin d th e en d abutment s o r retaining walls o f
integral bridges . Th e nex t stage s o f th e testin g woul d examin e changin g th e
important parameters . Th e effect s o f cree p were suggeste d a s likely to ameliorate
the exten t t o whic h bendin g moment s woul d increas e wit h increasin g abutmen t
stiffness.
Mr Lo w emphasize d th e considerabl e benefit s provide d b y th e horizonta l
stiffness o f th e highwa y pavemen t whic h coul d doubl e th e stiffnes s provide d b y th e
abutment. I t would be vulnerable t o any trenching carried out acros s the pavemen t
adjacent t o a n abutment . H e suggeste d a diagrammati c relationshi p whic h
demonstrated the interaction between th e force in the deck and the displacement a t
each abutmen t (Fig . 4in th e paper) .
Dr Thorbur n described th e problem s behin d th e abutmen t when usin g sof t
clay fills. Embankment s are usually constructe d in stage s with o r without provisio n
to spee d th e dissipatio n o f por e pressures . Bridge s ar e comparativel y rigid
structures. Th e problem s occu r from the need t o backfill quickl y between a bridge
and a n embankmen t whic h usuall y ha s differen t settlemen t o r consolidatio n
characteristics. I t is important to carr y a run-on slab well back from the abutment .
There seeme d t o be agreemen t that replacing the abutment with a pier and carrying
the dec k t o a ban k sea t fro m which a run-o n slab would ac t would overcom e suc h
problems. Alternativel y the wedge o f filling behind a n abutment would be stiffene d
by th e inclusio n o f sof t pile s under a run-o n slab.
Bridge deck s an d pavements shoul d ideall y hav e simila r temperature-related
characteristics. M r Low emphasized th e advantage s in thi s respec t from the use o f
lightweight concret e for integra l bridges.
CONCLUDING SUMMAR Y
A.M. LOW
The Chairma n starte d th e discussio n b y offerin g som e view s o f a contractor , a

viewpoint whic h ha d not bee n represente d in th e presentations:
Bridges mus t be maintainabl e and the requirements for maintenance shoul d

be include d i n th e basic concept.
Bridges must be saf e to build an d formal quality management systems for the
design ha d a n important role to pla y in achievin g thi s safety .
Any critica l requirements for constructio n should b e clearl y stated from th e
start i n th e documentatio n o f th e design .
To ope n u p th e discussio n h e aske d i f ther e were an y deservin g topic s whic h ha d

been omitte d thu s far . I t was suggeste d tha t slender column s migh t no t hav e bee n
covered. A trul y joint-free bridg e would als o hav e n o joints betwee n th e dec k and
the column s whic h woul d b e monolithic . Thi s i s a commo n desig n featur e whic h
results i n column s whic h ar e slende r t o minimis e th e stresse s du e t o temperature
changes. Th e desig n o f suc h column s require s carefu l though t becaus e o f th e
contradictory requirement s for vertical strength and swa y flexibility . Curren t rule s
can b e undul y conservative . Wha t is neede d i s clea r guidance fro m someon e wh o
fully understands the background to th e various code rule s for slende r columns .
Another topi c whic h ha d no t bee n covere d wa s fatigue . Dec k continuit y
introduces reversal s int o th e moment s a s a vehicl e passe s an d thi s ca n greatl y
increase th e fatigu e damage .
The Chairma n als o aske d fo r topic s whic h require d furthe r work . Th e
inter-relationship between crac k control, waterproofing and durability was identifie d
as a topic needing furthe r study, particularly for the slab s o f composite deck s whic h
had been mad e continuous . I n the hoggin g region s the y were in axia l tension ove r
their full depth .
A reques t wa s mad e fo r mor e desig n studie s o f joint-fre e bridge s t o b e
published t o encourag e designers . I t wa s reporte d tha t th e Stee l Constructio n
Institute would b e lookin g int o this .
There was also a request for more money to be availabl e for instrumentation,
both o n full-scal e bridge s an d o n th e centrifug e model s i n Cambridg e which ha d
been describe d i n th e previou s session . Thi s le d t o furthe r discussio n o f th e
centrifuge experiments . On e participan t felt tha t it shoul d mode l fill s which wer e
more compressible . Anothe r took the opposit e view that it should mode l th e stiffe r
290 Low
response o f cemente d fills .
Final word s
The Chairman' s final word s were "than k you". H e thanke d al l thos e whos e effort s
had mad e th e colloquiu m happen .
Brian Pritchar d the n close d th e proceedings . H e wa s please d tha t th e
Henderson Colloquiu m ha d maintaine d it s internationa l involvement . Th e
participants had come from eight countries . H e considere d tha t the presentation of
the papers had been a success, covering very many aspects of the topic. H e als o felt
that th e discussion s ha d bee n successfu l i n buildin g o n th e contribution s o f th e
papers. H e announce d that , following th e usua l practic e for Henderson Colloquia ,
an open meetin g would be hel d a t the Institution o f Structural Engineers in London
to repor t th e proceeding s an d to allow thos e no t presen t to ad d their contributions.
He ende d wit h som e reflection s o n ho w th e continuin g developmen t o f
technology woul d affec t th e desig n o f bridges . H e ha d recentl y visite d German y
where h e ha d bee n introduce d t o th e "intelligen t bridge " which monitore d it s ow n
strains an d corrosio n and he wondere d i f thi s would become standar d practice.
A U T H O R INDE X
Bier, G. , 18 3 Nicholson, B . A, 3 3
Burke, M . P. , 21 3 Norrish, A R . M. , 25 1
Clark, G . W. , 19 3 Owens, G . W. , 16 7
Cuninghame, J . R., 1 1
Parsons, J . D., 8 5
Dumont, R. , Paterson, G . P. , 16 1
Pritchard, B . P. , 55 , 203 , 22 9
Fuchs, F. , 17 7
Quinion, D . W. , 28 7
Hambly, E . C , 16 7
Shanmugam, S. , 72 , 17 3
Johnson, I . D., 3 Shioi, Y. , 7 5
Springman, S . M. , 25 1
Kumar, A. , 23 9
Takano, H. , 20 7
Low, A . M. , 41 , 265 , 28 9 Taylor, H . P . J. , 6 5
Lund, J. , 2 1 Thompson, D . R. , 19 3
Thorburn, S. , 27 3
Martinez y Cabrera , F., 12 9 Tilly, G . P. , 4 5 , 9 3 , 2 4 7
Muller, J. , 10 5
Veje, E . M., 9 7
Nakamura, S. , 7 5
S U B J E C T INDE X
Abutments 167 , 213, 251, 26 5 Composite constructio n 10 5

Access problem s 2 1 Concealed drainag e 22 9
Active contro l 10 5 Connecting sla b 23 9
Aesthetics 8 5 Construction aspect s 21 3
Anchor lengt h 18 3 Construction dept h 5 5
Arch bridges 4 5 Construction joints 6 5
Arches 19 3 Construction system s 12 9
Continuity 4 5
Asphaltic plug joints 1 1 advantages 5 5
effect o n bearing s 2 1
Backfill 21 3 plate 23 9
Ballast, railwa y track 18 3 precast beams 6 5
Bangkok Expresswa y system 10 5 systems 12 9
Beam continuit y 22 9 Continuous arche s 19 3
Bearing fault s 2 1 Continuous composit e bridge s 23 9
Bearing load s 2 1 Continuous concret e slab bridges 21 3
Bearing movement s 2 1 Continuous girde r bridges 7 5
Bearing reductio n 55 , 65 Continuous welded rai l 18 3
Bearing stiffnes s 18 3 Continuously reinforce d concrete
Bearing type s 2 1 pavements 17 7
Bi-stayed bridges 10 5 Creep, integra l bridges 3 3
Box girder s 97 , 129 Cycle contro l joints 21 3
Bridge sea t joints 21 3
Buried structure s 16 1 Dehumidification syste m 9 7
Cable profil e 12 9 Danish continuou s bridge s 9 7
Cable-stayed bridges 10 5 Deck constructio n depth 5 5
Caselle Viaduct , Italy 12 9 Deck length s 18 3
Centrifuge modellin g 25 1 Deck sla b continuit y 55 , 229, 239
Clarea Viaduct , Italy 12 9 Deck sla b hingeing 22 9
Clays 27 3 Department o f Transpor t advic e not e
Closed bo x girders 9 7 161
Comb joints 7 5 Design aspects , semi-integral bridges
Combined-type jointless method s 20 7 213
Composite bridge s 85 , 239 Design, continuou s viaduct s 12 9
294 Subject index
Design, precas t beams 6 5 James Rive r Bridge, Virginia, US A

Detailing o f joints 3 105
Durability 16 1 Japanese bridge s 20 7
composite bridge s 8 5 Joint type s 11 , 207
Jointless viaduct s 16 7
Ebron Crossing , Franc e 10 5
Economic aspect s 21 , 55, 85 Large scal e testin g 1 1
Elastomeric bearings 2 1 Lateral restrain t 21 3
Embankments 265 , 273 Limit state s 6 5
Embedded joints 20 7 Locally separate d deck sla b 23 9
Embedment, precas t beams 22 9 Longitudinal restrain t 21 3
Expansion joints 45 , 97, 18 3 Long-span girde r bridges 10 5
performance 3
Maintenance, stee l viaduct s 16 7
Fabreeka bearing s 2 1 Masonry arc h bridges 4 5
Far0 bridge s 9 7 Masonry arche s 19 3
Fatigue 16 7 Material tests , joints 1 1
Fatigue testing , joint s 1 1 Metal bearings 2 1
Finite elemen t analysi s 26 5 Mohr-Coulomb effect s 26 5
Flexible pier s 8 5 Monterrey rai l viaduct, Mexic o 10 5
Flyover lengt h 5 5 Movable joint s 21 3
Foundation saving s 5 5 Multi-span composit e bridge s 23 9
Four-span bridge desig n 3 3
Neches Rive r Bridge, Texas, USA 10 5
Geotechnical behaviour , embankment s Non-linearity 25 1
273
Girder desig n 16 7 Oleron Viaduct , Franc e 10 5
Granuda Viaduct , Ital y 12 9 Overpasses 17 7
Great Bel t Lin k 9 7
Pavements 177 , 265
Hysteresis, abutment s 25 1 Pier desig n 16 7
Historical revie w 4 5 Pier saving s 5 5
Precast bridg e beam s 23 9
Instrumentation, embankment s 27 3 Precast pretensione d concret e beam s
Integral abutment s 85 , 251 229
Integral bridges, temperatur e effect s Precast segmenta l constructio n 10 5
33 Prestressed beams, desig n 3 3
Integral crosshead s 22 9 PTFE/steel interface s 2 1
Integral stee l crosshead s 8 5
Interaction, bridge s - embankment s Rail expansio n joint s 18 3
273 Railway bridges 18 3
Iron bridges 4 5 Railway viaducts 105 , 167
Isere Rive r Bridge, France 10 5 Reduction i n dept h 5 5
Isolation plat e 23 9 Reduction o f joints 5 5
Italian joint-free bridge s 12 9 Reinforced concret e 17 7
Subject index 29 5
Research requirements , joints 1 1 Temperature cycle s 16 7

Restello Viaduct , Italy 12 9 Temperature effect s 33 , 45, 183
Retrofit continuit y 8 5 abutments 25 1
Roize Bridge , France 10 5 Temperature movement s 26 5
Rotational restrain t 21 3 Testing, larg e scale 1 1
Rubber bearing s 7 5 Thermal fatigu e 16 7
Run-on slab s 161 , 273 Top sla b continuit y 23 9
Traction effect s 18 3
San Antonio Y project , Texas, USA Trans-Tokyo Bay Bridge 7 5
105 Type approva l tests 1 1
Segmental constructio n 10 5
Seismic desig n 7 5 Vertical restrain t 21 3
Selection o f expansio n joints 3 Viaducts 19 3
Service performance , joints 1 1 Windward viaduct, Hawaii 10 5
Settlement o f embankment s 27 3
Shell section s 12 9
Shrinkage, integra l bridges 3 3
Single spa n porta l 3 3
Site influenc e 3
Soil-structure interactio n 25 1
Space trusse s 10 5
Specification o f joints 3
Stability o f embankment s 27 3
Standards 16 1
Steel beam s 23 9
Steel bridge s 8 5
Steel composit e pier s 7 5
Steel viaduct s 16 7
Strain cyclin g 25 1
Structural analysi s 12 9
Superstructure restrain t 21 3
Surfacing 17 7
Surveys, joints 1 1

Continuous and Integral Bridges by B. Pritchard

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Continuous and Integral Bridges by B. Pritchard

Uploaded by

Copyright:

Available Formats

Continuous

Bridge Bearings and Expansion Joints

Pembroke College, Cambridge, UK

Taylor & Francis

First editio n 199 4

Transferred to Digital Printin g 2 0 0 6

© 199 4 Taylor & Francis

Apart fro m an y fai r dealing fo r th e purpose s o f researc h or private study , o r

A Catalogu e recor d for thi s boo k i s availabl e fro m th e Britis h Librar y

Library o f Congres s Cataloging-in-Publicatio n Dat a availabl e

Contributors and participants i x

PART ON E PROBLEM S O F BRIDGE ARTICULATION 1

1 Performanc e of deck expansion joint s 3

2 Researc h requirements for bridge deck expansion joint s 1 1

4 Effect s o f temperature , shrinkage an d cree p o n integra l

Report on Session One 4 1

PART TW O ADVANTAGE S O F BRIDGE CONTINUITY 4 3

5 Historica l revie w o f th e developmen t o f continuit y an d

6 Advantage s o f bridge deck continuity 5 5

7 Continuit y i n decks with precast beams - practica l issues 6 5

PART THRE E CONTINUOU S BRIDGE S

8 Desig n concept of the Trans-Toky o Ba y Bridg e

9 Continuou s steel and composit e bridges

Report o n Sessio n Thre e

PART FOU R INTEGRA L BRIDGE S

10 Danis h integral bridges

11 Precas t segmental constructio n

12 Italia n joint-free bridge decks

13 Th e U K Departmen t o f Transpor t vie w o n continuity /

14 Jointles s steel viaducts for railwa y

Report o n Sessio n Four

PART FIV E LONG-LENGT H CONTINUIT Y

15 Th e desig n an d performanc e o f continuousl y reinforce d

16 Experienc e i n th e us e o f continuou s welde d rail s o n

Report on Session Five

18 Stud y on jointless bridg e construction in Japa n

19 Semi-integra l bridges : a concept whose time has come?

PART SEVE N PRECAS T BEA M DEC K CONTINUITY

2 0 Continuit y o f precas t pretensione d concret e beam s b y

21 Locall y separate d dec k sla b continuit y i n composit e

Report on Session Seven

PART EIGH T SOIL-STRUCTUR E INTERACTION

22 Soil-structur e interaction : centrifug e modellin g o f

23 Th e contributio n o f th e roa d pavemen t i n absorbin g

Report on Session Eight 28 7

Welcome t o th e proceeding s o f th e Britis h Grou p o f IABSE' s annua l

The late Bernard Godfrey, fondly remembered as a very active member of

long. I t is als o eviden t tha t longer an d longe r integra l bridges ar e being

Continuous and Integral Bridges. Edite d by B.P. Pritchard. Published i n 1994

2.1 Performance requirements

2.3 Influence of site on joint type

deterioration of the adjacent surfacing and the eventual potholing of

performance requirements for expansion joints, eg. the transfer of

3 Detailing and specifications

The selection of an appropriate joint must be supported by the

drainage can however be alleviated by the use of buried joints if

This paper is based on work conducted as part of the research programme

Booth, Sweetman and Woolfe Consulting Engineers and Architects (1987).

Price, A.R. (1984) The Performance In-service of Bridge Deck Expansion

Bridge substructure s mov e ver y slowly , bu t th e dec k i s

maintained, an d replace d severa l time s durin g th e lif e o f

• T o provid e a sea l betwee n tw o part s o f th e structur e

Several type s o f expansio n join t ar e use d i n the U K t o

3 Review of Department of Transport (DOT) research