You are on page 1of 22

B ridges w ith M ultiple C able-stayed Spans

M ichelV irlogeux,CivilEng.
Consulting Engineerand Designer.Bonnelles.France

Sum m arl' Brooklyn B ridge and G arabit


Viaduct.
This paper is devoted to a new field of application of cable-stayed bridges: Furtherm ore. the Lake s' laracaibo
bridgeswith multiplecable-staqed spans.ltbeginswith ashorthistoricalrevieqv Bridge has som e com m on pointsw ith
and aspecialreferenceto R iccardo slorandi-sbridges.ltthen describesthe N'ery the bridge over the Firth of Forth.
few bridgesthathave been builtrecently with multiple cable-stayed spans and Both were technical deadlocks.new
thedesignsthathavebeen proposed in the lastthirtj'years.ltendsw iththepre- structural concepts which were im -
sentationofthreerecentmajorprojects.atleastoneofwhichwillbebuilt.These m ediately surpassed by m ore efficient
three projectsshow the extrcme efficiency ofthisconceptfor crossing q'
ide ones: classical cable-stayed bridges.
riversand sea channels. u'ith tlexible towers and continuous
decks-condemned N' lorandi-s concept
in the same wal'thatclassicalsuspen-
sion bridges condem ned the sophisti-
H istoricalR eview A lbert C aquot erectcd the cable- cated truss structures inspired by
stayed bridge overthe Donzère Canal
Fowlar and Baker's Bridgtt.But.like
in reinforced concrete. This bridge. the èridge os'erthe Firth ofForth.the
The first cable-stayed bridges were x'ery often forgotten in the history of
erected at the beginning of the 19th Lake M aracaibo Bridge is adm ired by
cable-stayed bridges. preceded by
century- but disasters occurred with architectswho understand the evident
som e years the fam ous Stromsund flow offorcesand who aresensitiveto
the collapse of 1he bridges over the
Bridge erected in Sweden under the impression of strength that em -
Tu'eed and Saale rivers.Designers at D ischinger'sinfluence.A fantasticde-
the tim e ignored the real flow of anatesfrom them assand shapesofthe
velopmentofcable-stal'ed bridgesfol- structure.
forces.did not consider wind effects. low ed.firstin G erm any and later a11
even through an evaluation of static
overthew orld. N evertheless-R iccardo Nlorandibuilt
forces produced by wind pressure on
the structure. The French scientist But justwhen the design ofmodern severalbridgesaccording to the same
H enriN avier es'en ''proq'ed''thatsus- cable-stayed bridges w' asbeing devel- principles:
pension bridgesare farm ore efficient oped.with t -lexible tow ers and a con- -
tinuous deck.Riccardo Nlorandiori- the PolceN'era viaduct near G enoa.
than cable-stayed bridges.' These acci- Italy.built between l960 and 1964-
dents and this scientific dem onstra- ented hisown ideasin a very different
with three towers and two main ca-
tionshalted the developm entofcable- direction.H is towers w ere extrem ely ble-stayed spans each 280 m long
stayed bridges for almost a century. rigid.in the shape ofa portal fram e
and cablestayswereonlyused in som e longitudinally (an inverted V).within- - (Fjg... /)
tw o sm allerbridgesin ltaly.i.e.the
suspension bridges close to pqlons in clined strtlts to support the deck ata bridge overthe Tevere ats' Iogliana
orderto stiffen the structure:the m ost distance on each side.Each towersus-
built between 1963 and 1967.and
famous exam ple being the Brooklyn pended 8 1 double cantilever - w ith a
rigid connection to the tow er - and theCarpineto Viaductbuiltbetu'een
Bridge.USA. 1971 and 1974.w ith one and tw o
drop-in spans produced the link be- t05VCI-S.reSPeCtiVely
In France in the firstyearsofthe 2()th
century.A lbertG isclard erected sever- tweenadjacentcantiles'ers. - the 'Bs
adiKufBridge in Libya.btlilt
albridgesB' ith a specificeable system . Thisprinciple wasapplied forthe first between l965and 197l.u'ithtw o py-
interm ediate between suspension and tim e for the erection of the Lake lons and a Iuain span of 28l nR.
cable staying.giving m uch m ore im - Nlaracaibo Bridge in Venezuela (Figs. which u' as for seven years the
portance to the cable stays.H is ideas ,/
--?). which was built betvveen 1957 longest concrete cable-stayed span
were occasionally reproduced. and 1962 gl1.Thecentralpartofitis in thew orld.
madeofsixtowersand fivem ain spans O nly one bridge hasbeen builtby an-
However.itwasin Spain in 1925 that each 235 m long.The erection ofthis
Eduardo Torroja built the firstreal otherengineer follow ing these princi-
cable-stayed bridge in concretc. the bridgewasa maiortechnicalachieve- ples. nam ell' the Chaco Corrientes
m entatthe tim e.and can becom pared
Tem pulA queduct.Here.a cal7le stay Bridge over the R io Parana in A r-
to the construction ofthe bridges be- gentina. designed by Jean Courbon.
sim ply replaces:1classicalsupportthat tween H onshu and Shikoku in Japan.
could nothave been builtdue to the or of the Storebelt and O eresund This bridge was com pleted in 1973
siteconfiguration. w ith two pylonsand a centralspan of
Bridges in N orthern Europe.For lhis
The real developm ent of cablu-stayed reason. the Lake Nlaracaibo Bridge 245l n (Fig.-
S).
bridgescam e with the ideasand publi- deserNesto bepartofthe seriesofthe Before concentrating t)n otlr them e.
cations of Franz D ischinger in the m ost fam ous bridges over the world. wernay evoke SomemaioraspectsOf
1930s and 194()s.Surprisingly.the first u'ith the G olden G ate Bridge- the the evolution of m odern cable-stayed
application NvasinFrance in 1952when bridge oN'er the Firth of Forth- tht? bridges.

StructuralEngineering lnternational 1/2001


' A majorstep cannewiththeI
Rotionof
23.%n-l 23'
7rI'n -'RSl'n 27,<kn'k 235nn 23
.'
I'1'1'1 n-lu1tip1e c:1ble stays suspending t11e
') lyt?nl 4 '1 --
I
1 jj deck at close inters' als.developetlby
.
1 . . rjy
', 1 p1 .
Helnatlt H ol-nberg for the erection of
the Friedl -ich Ebert Bridge in Bonn.
- -- - - '
- .. .-w Y. = - '......... .- . ......
.... m * '
w = '= - - - - -
Germany.in 1967.A furtherstep was
the concept of cable-stayed bridges
with flexible decks.inN'ented b),
't.
-Irich
Finsterwaldel-andFritzLeonhardtand
deA-eloped b)' Rend ' h.'
aIthe1-f0r the
erection of the Diepoldsau Bl-idge.
'
Switzerland. in 1985 and by Jörg
::,4
...
' 1 Schlaich u'ith the Eqrripos Bridge.
I I
$
br
r1
.j
' Greece.colupletedin 1993-aprojectin
l
which the author took partast2o1 )sul-
tantforthe G reek attluinistratitln.
Rapid prtlgresshasbeen I-nadt?il1span
length in recent j'ears. and cable-
stayed bl-idgesnow con-lpete with sus-
pension bridges1-
01-spansbetvvet'n 70()
and 1200.oreven l5()()m.
Theerectionofthe N orm andie Bridge.
France.wasamajorstep in thisfield.
The author desveloped the design nt
the Sers'ictt d'études techniques des
routesetautoroutes(SETRA )in asso-
ciation u'
ith seN'eraldesign officesand
laboratories (Sofresid-Sogelerg.Seee.
O uadric.Onera and CSTB).takil' lg in-
spiration frol'n the lectures of Fritz
Leonhardt- who stated long agk)that
it is possible to erecl cable-stltj' ed
bridgesw'ith spanstlp to 1500 nn- and
Fiy.2:1-
.
'?
-(
.
a('
?à()??()ff/pc1.
-tl&'
t.
ahlaracaibobé'idge fron'
l Englisll suspension bridges ftlr
their strealulining.The execution de-
sign '
was developed 1* 4.
)1- the concrete
partsofthebridgebythe GlE.du Pont
de N ornxandie (incltlding seven n- ta-
ior contractors: Canlpenon Bel-nard.
Bouygtles.Sogea.GTNI.Dtlmez-O tlil-
lery and Spie BatigntllIes)and lklrthe
steelpartofthe centralspan b)'N10n-
berg and -l-horsen u'ith the participa-
tion ofCOSVIConsult.BilfingerBerg-
erand FrcqrssinetNvere subcontractors
for the l-oundations and for the pre-
stressingand cablestaq,ing.respectiN'
ely.
The N'ornRandie Bridge is no longer
the longest cable-stal'ed span in the
u' orld. Since N'laq' l999. the world
record belongs to thtt Tatara B ridge.
Japan-:1N'ery elegantstructure.l11the
author-s opinion-this is the nzoste1e-
gantofa11Japanese bridges-showing-
i1'itwere necessarl'- thatcable-stayed
solutions can be both extrenèel)'effi-
cientand elegant.
The design of long-spëtn cable-staq' ed
bridgesisdoluinated bl'the resistance
to turbulentw ind dl'naillic effects.itnd
by aerodq'nannic stability'
.Streallllink?d
box girders-inspired by the English
suspensionbridgesand theN 'orn'1andie
Bridge. constitute the best technical
Repklrts StructuralEngineering International 1/2001
'

- I j
l
#j t
.
Q). , - t
#
. .. p. -
. l. x.
, < j.. .'
*.
- ... . A N *.*.ehw
... . . j. ! -. +z t
. ->*'
' .

. -. .. .x - . .% ..>
.g.
.. . . . .t
r.
>...-... v.. --.o -. .- . . =. w . . .s-.= J..--g.
% 6.
.* .o: . z>. .. -.....-w. .v.r.....

/:ïy
. (7
'
..3
-.'.
TI1(.
7(L
'
IIf'
It
r
.(
''
?(l
'
t
.)''
F7'
t
??lft
?-
'
f1)Ftt'
/>
c.''62/,'?/2(3d
-t'#rt.
-!.
%sil'
.t;;
./
'

solutions to these problems.The au- quenciesofthe structure itself.ltises- um in the applicationsofprestressing


thorconsidersthattheyalso havetobe sentialthatthese aiguilleshave a high stranbs.from internaltendonsto cable
Preferred for shorter spans. even internaldam ping.in addition to ahigh stays. passing through external ten-
though they are slightly more expen- fatigue resistance- in order to con- dons and extradosed cables.and per-
sis'
e.because they behav'e better than tribute to the globaldam ping and to haps ending with suspension cables
deckswith l-girders.The latterhaq'eto avoid :1 transfer of energy to other m ade ofparallelprestressing strands.
bc eqttipped with appendices.such as typesofvibrations.In sum luary.aigu-
t'
airings and baft
les-in orderto reach illes are n-
lore adapted to long orvery
SpeciscProblem sofBridges
an acceptable aerodynalnic behaviour. longspans.
which they do not have naturally with M ultiple Cable-stayed
through theirshapes. Nluch progress has been m ade in the Spans
design ofcable staysthem selq'es.pass-
Cablevibrationshave affected several ing from lock-coil cables to parallel ln a classicalthree-span cable-staq'ed
maior bridges in the last ten years. wirtzsundt?rthe influence ofBBR and bridge. Ioading the m ain span pro-
D espite better understanding of the
FritzLeonhardt.and m ore recently to ducesadow nw ard deflection and.due
phenomena that produce such vibra- cables m ade of parallel prestressing to 81 tension increase in the corre-
tions.one cannotconsidertheproblem strands.The laststep in this direction sponding cable stays. a detlection of
solved.assom epointsare stillcontro- was the developm ent in 1988 of both pylons towards the loaded span
versial. Som e cable vibrations naight parallel auto-protected strands by
be produced by lim ited m oN'em entsof (Fig.6).The cable stays that suspend
Freyssinet.s'lodern cable stays m ade the side spanssufferonly lim ited ten-
a deck that is not w ellstrealnlined: sion variations-and side spans deflect
m ovem entsw hich induce:1param etric of parallelstrands provide very good
proteclion against corrosion. high upu'ardsin theglobaldeform ation due
excitation ofcablestal' s.Rain.in some to their reduced rigidity. O nly back'
-
strength.and high fatigue resistance.
cases.m ight accelerate the phenome- H owever.ifalldetailsand the need of stays.which are anchored close to the
non to :111alarm ingscale. dom inating cable vibrations are con- abutments.are subiected to high ten-
O n the otherhand.itisknown how to sidered. then we have to speak of sion variations since their lou'er an-
master cable vibrations by different eable-staying systems rather than of choragesare alnnostfixed and cannot
types of counterm easures, To elim i- cable stays. Existing specifications give u'ay in the sam e naannerasother
nate rain-and-w ind induced N'ibrations. about cable stays. which were not ones in side spans.These back stays
ductscan beshaped tochannelrainwa- based on a very scientific approach. controlthe pylon detlection towards
terdou'n the cables(asused on theHi- haq'e to be updated to adapt them to the loaded span.balancing practicalll'
gashi Kobe Bridge-Japan).A better realgoalsand needs. all the horizontal con- lponent of the
solution consistsofdestroying the co- tension variationsin the cable stal'sof
The tlse ofrod barsto constitutecable
herence of excitations by installing staq's has not been evoked.as tlley are the nlain span.
thin helicalfilets on thtzducts.which considered notadapted to such :1use. The disyn- ll-
netry in the distribution of
do not increase the drag forces very due to theirlowerductility.theirsensi- tension N'ariations in the rear cables -
much (as used oIz the Normandie tivity to bending stresses. and their high tension variation in back stays
Bridge)-orb)'creatinga seriesofdim - and N'er).
'low tension variation in all
plesin the ducts.u'ith :1random distri- lowerfatigue resistance.
ot11e1-cablk? stays in t1-
1e side spaIAs -
bution (asused on theTatara Bridge). T'he need form ore m odern specifica-
produceshigh bendingn' lol-
nentsin py-
A notherapproach.adapted to alm ost tions is enhanced by a new develop- 1ons.In orderto reduce thenn-itis nec-
alltypesofcableN'ibrations.consistsof m ent of cable-stayed bridges con- eSSa1'y to Concentrate anchorages in
increasing the dam ping in cable stays ceived by Jacques Nlathivat. nam ely
the pylon heads in bridges that haN'e
by the installation ofdam pers ofdif- extradosed bridges. which becam e such : 1configuration.The reduction of
lkrenttypesatthe loweranchorage to very popularin Japan.Som e engineers the distancesbetNveen the attachm ents
the deck.or at both anchorages.To use the sam e specïfications for extra-
ofrearcablestaysreducesthe bending
elim inate param etric uxcitation. the dosed cables as for prestressing ten- m om ents produced by the concentra-
bestsolution generallq'is to attach al1 dons since they are m uch m ore
tion oftt?nsion N'ariations in the back
thecable staysin a planew ith aiguilles faN'ourable than those for cable stays.
(cross cables) to change the natural EA'en though an interm ediate stress stays.
frequencies of cables in their plane level has been adopted in Japan for B''hen :1side sp84n is loaded.it detlects
(verticalvibrations)andtomakethem extradosed cables. it w ill becom e logically doïvnNvardssvith a tensik)l' 1in-
very different from the natural fre- necessaryto deN' elop akindofcontinu- crease in a1lcable stays thCttsuspend

SlructuralEngineering International 172001


H'
the loaded span.except for the back havesas ifthe side spanswere totally passes directly to the internAediate
stays.'
Fhe tension increase in the cable rigid and the lower anchorages of all piers.Thus the cable stays anchorcd
staysprodtlcesadeflectionoftheadia- the cable staysfixed.In thissittlation. close to the abutlnents receive stress
cent pylon towards the loaded span. the pylon detlection toss'
ardsthe main variations that are only half or one
This restllts in 8
4 rcduction of the ten- span is reduced and the efficieno'of third ofwhatthel'would bein thepre-
sion in the back stays and an upward the cable-staying system increased. vioustypeofcable-stayed bridges.
deflection ofthe main span. Even the doNvnward detlection ofthe
main span isreduced by the reduction Very sim ilar results are obtained with
Back stays evidently have a specific a cable-stayed bridge having a single
role in the structuralbehaviourofthis of the pylon deflection.The indirect
butdecisive advantage ofthissolution pq'lon when interm ediate supportsare
type of bridge.and suffer the higher installed in the shorter span.i.e. the
stress variations from Nvhich fatigue is thatanchoragesofcable stays can
be distributed overa rather large dis- side span.
m ightdevelop.
tancein thepylons.withoutproducing ln contrast.the behaviour ofa bridge
The sam e type o1'behaviour can be high bending m omentsbecause ofthe with m ultiple cable-stayed spans is
seen in three-span cable-stayed bridges presence of an efficient back stay in cornpletely dit-
ferent.Itishighly infltI-
thatare extended on each side by an frontofeach cable stay com ing from enced by the type ofconnections be-
accessviaduct.Thecablestaysthatare them ain span. tween thedeck.piersand pylons(piers
anchored close to the end supportof below thedeckand pylonsaboq' eB'hen
the cable-stayed section behave as ln addition.whena side span isloaded they'are independent m em bers. be-
back stays.The structuralbehaviouris there is practically no effect on the com ing towersNvhen they constitute a
also the sam e in bridges vs'ith a single cable-staying systcm. all the load single.continuottsstructuralnzember).
Pylon and tw'o spans. if one of the
spansisslightly shorterwith back stays '

anchored closeto theabutment. >'


w)
.. .xcn-r .. ! .:
5 X>''
.e A
. .... .-
A classical technique to im prove the . ' /
h
structuralbehaviourofa cable-stayed
bridge is to proNride interm ediate êl)S!a!1ccl''n!'lt!uratlon
supports in side spans (Fig.7). 7Yis
conceptwasdeveloped rapidly in Ger- -.
m anyand wasreproduced 1- 01-the erec- Nyyyyy /1
j'J '
tion ofthe SeysselBridge-France.and
latertheN orm andie Bridge.Svhen the
. '
-
.
K J
m ain span isIoaded.al1cable staysan- bJLoadingitccnlraispan
chored in side spans behave as back
stays.This is true not only for those
that are anchored close 1() the abut- .- .--- h'v. 4444,4 1
/
, .-.
. -
'
. ,sw
X ''
'---w
m entor to the interm ediate piers.but
also forallotherssince theshortspans
between interm ediate piersreducetl ze
c)Load1nganadjacenl<pan
upu' ard deflections in side spans to
N'eo '1ow levels.The structure thus be-

Reports StructuralEngineering International 1/2001


thecable staysare concentrated into a
singlestrutatthecantileverends.Nvith
a untformly distributed cable-staying
='
Je-.
-.
'> x
system -asdeveloped in m odern eable-
4sNe
u.x
4v
.-
e
.,.v:..:t<t-r.=xo.
.r&-r-d'
# 5#.x#-..
c.-!. %- -.'
-.x..?#-..o. -1/
.a-v
'Nuw
-.s*'='q;wz
k-1 '-=.e=-<=..c=.#M'=N=.z
Jre A.'
<xw.5z
'
v -ex*.x1x..-
e x--
'w
+p
.x
,,=*=-
:&. .>'us
-=.rv ..w.sr
'=v> ''
e
.-r
-;
-r
.nznd
' T7
'
,x -A
-.'='.v
&.
w
.'=m-
-T&.
ëe
''
='-
:'e5
-.r.>
- *-'
v'.=
A+
.m=:-5-,--
*'x..a<
, =.
'+I?.
E.
x'J
'>
><-7='..:-.
=2 -<(
<.=
,.
-en.=*
J/ =.=
2. ':
.r5
.'
=%x
M-'-4te*
. J=>z.
'
,.
+ .xj!.. %:
.<=:. L= .-xsrg-J'
.. t-.:
>.w+$.<: ..+<z.....%+. .-R:>
..<...-..o. =w=....>..x.a'Qu.:'.:
=..<%'.mîu-I:r..
xa.c.r''-+.
*v->W stal'
ed bridges.theseinclined strtltsare
a)Sta:1'-'
-()nfi&
e'Ju-'3t1t)n notneeded.A nd the drop-in spansbe-
tw een the cantilevers allow for free
length variationsin the deck.

The drawbacks of this structural sys-


'->s:'4#t-t-+uùv!N2--'>rO>oCS.;A.'/A.,o...i...ex..:.e'.:5l4e.ccrêxo.e'
u ezwuv-.zsgz>..,,v.;!-s?.L.A7uq.-s.ArDr-.paz.av4rïrPsDwzL.zx-tvste-.,
sx..
x :z,%a..zc.>s.%'..
a&->ZJAJ/-a.:)/.p
%r-a.'JR' fW'
?'7- ,ler%vzw..rztt?r
lt -Jcrqït.+nr.-.4.J-twx.k<.auA'%n>
.tv,auvvz@c:.z:r:tl);.mquè'
Ji>i,..
p7Jai.yrzkei.Xx
v..<p'
lD2ru-p'.e.r'.i7.-T.N>1's!4=rAAi>t5r4vJ'->.axs.
: .:.
a*.
s +7,.>.:MïYTa''er.fxv<z1!ti.gk'hv-
.b
-c4.k,*.
:keç tem are its high cost(induced b)'the
lo)Effeclt)n t 'ts17t'tn l()ad1no ' w1t1- 11)s1m plc l1nc t' l1'beari1'1ulso1. 1tht 2p1tlrs :t)sur7p6.prltht.tt'y>tktl' n tà1'uiec' K and p'$'lon>. structuralcomplexity and them ultipli-
no part1c1pi 'tt1t)nk11't11e p1ers.a1:tl1 f7h .' l(.. )1'1sturn111u u: al111k''' k'tfree1' .$
cation of the construction equipm ent
for diffurenterection techniques).the
large N'olum e of concrete needed for
the differentstructuralmembers.and
1h
,e --
r..
. A..-.
4-t
. l.L$
- ...f.5è,
-c 'e!
rr
<A
zfgxif-
'v.
..n.
a7,
o'.
t
o
' 1gvq.
irr-
.
,
yv
q*
.
r-ik
Ec
. l.
-.
k
.
Jz
sz4s.
'x
c -.
.
%
-,=
..
%
aA
.,-Si
J -P
zgl
. xé lsx4
-:v
' 4r4
,x..
l v.
a. =*-
.
.sîarl -.
r..t:J
'2'z,z-*
kw .x-
$--: :s;
ga. 3#
-.p
:4-
;6
z.g'. .-
y.. &zr.
?< q.
ja=svl.I-
=J
' g:-
-
. A.@c%t
.
.. :ar-
.
.
#
(-
.u.-
,-
u*r.
.ca?
'l
f.5
2.bv
- ïv%.-e
r.7
z*:w/
Rx$
v
:a
'7X-%!!t
. zd.vex
ï rx
.Ln.ç:s
pg:o
><. -.-
--,
. -s
w
u
(ùl
s-,
;i /a
...- z,-
1;. r>*
c. rr
dp
s-,.
- -..%
a= ë..
A..t
k
..
-
r-i. x*-
1tu::er.*
.
,
.
i.
>
ikxk
.v.
->
.u').v'
W-F;.Xx
k-:.?
q.
fcz.
>
:h.
R
,w.
7qa.=-.
! dA
t. be.
r,%
.îllrr
z%
vu
-rl.--
'
4
'-.
w'sS
z-s'x
:
. k.%q
,a-:>
>I' the Iarge numberofexpansion ioints
J)F- 1'I
''e0tth!'u
'l1 'k'.
'tdl
':1f's.
&11.
' 1'
11<7c'i'lt-t1nuz.p.l;
')r'
stak114srpê.
trti1'
lt11el1n'
11tat1tè1:(A'1'u'qtl1ect1on< w hicl) lil-
nit the user com fort and re-
.g'kl1)l
quire frequent and costly l nainte-
n ance.

A sstated above.the Lake 5laracaibo


Bridgeisoneofthemaiorbridgesbtlilt
during the 20th centurl'
.and itsdesign
Thisstudl'beginsw ith a bridge haqr
ing cable-staq'ingsq'
stem hasa N'erl'lim ited w asadapted perfectll'to the technical
towers.with a design N'ery similar to int-luence on the controlofdeflections needsofthisproiectand to t11e erec-
those of the classical three-span and bending momenls.Cable staysare tion lechniquesand ideas ofthe time.
bridgesdiscussed before. only efficient to balance permanent '
W-e m ay think that this fantastic
loads.self-weightand equipnlent. achievem enthad such an intluence on
svhen a span is loadtzd- it deflects
downwards.tension isincreased in the These tw o exam ples dem onstrate the R iccardo Nlorandi that it has been
cable stayswhich suspend it.and due basicproblem ofbridgeswith m ultiple very difficultfor him to develop neu'
to this tension increase the adiacent cable-stayed spans:hoqvto controleffi- concepts.even w hen he had to build
tosvers deflect towards the loaded ciently.economicallq'and elegantlq'de- cable-stayed bridgesw ith three spans.
span.'Thisdeflection liftstheadjacent tlections and bending m om ents pro- forw hich hissystem w asunnecessarily
spanswhich m oveupwards.and dueto duced b), ' live loads? But they also complicated. Only in his last years
this movementin the adjacentspans show thatthere isa majorinterestin could he design m ore classicalcable-
the other tower in each of them de- takingads' antageofthepierrigidity. stayed bridges. more efficient and
flectsslightly in the opposite direction m ore in agreem ent with the interna-
The best solution is to design rather
(Fig.(
$).This globaldetlection is on1/: tionaltrends.
rigid towerswith the deck passing in-
controlled by the rigiditl'ofdeck and
side.or to produce a rigid connection Several designers were inspired' by
tovvers.There isno back-staq'ing effect through the deck betyN'een thepierand Riccardo 5.
to lim itdeflections and deform ations 10randito develop b1-idges
thepylon.A rigid connection between with l' ntlltiple cable-stayed spans-but
and to im provetheefficiencyoftheca- the deck and the towersincreases the none ofthem hasbee11erected.
ble-staying system . structuralefficiency.b)'the transferof
Ifoneoftheadjacentspansisloaded. partofthe t-lexuraleffectsto the tou'- In 1967. Ulrich Finsterwalder pro-
it now detlects downw ards and the ers.I-loweN'er-this raises a new prob- posed for the Great Belt crossing.
spansand towers deflectin the oppo- lem :the structuralsystem hasto adapt D enm ark.aseriesofspans350 m long.
site direction than in the previous to thedeck length variationsproduced ' l''
he deck u' asextremelyslender.being
Ioading case. RXis nneans that each by temperature. concrete creep and J 'ustaslabin reinforced andprestressed
structuraln)emberissub2jected to i1-
1- shrinkage. as well as the structural concrete.and wasconnected rigidly to
tense bending m oments,ln one direc- shortening produced by prestressing the very rigid tow ers.w ith an expan-
tion and the other.resulting in high forcesinstalled afterclosing the spans.
stressvaria(ions.
sionjointatmid-span.Thepierswere
d-ivided into two shafts longitkldinally
A n even w orse situation appears w ith to atlow for length variations in tlne
a structure made ofacontinuousdeck Riccardo ( Nlorandi's Solution deck.Therefore.itw as notnecessary
to which the pylons are rigidly con- tohaveanexpansionjointineachspan
nected.and which issim pll'supported 'T'he structural system deN'eloped by ('Fig.I()).
by the pierswith a singld line ofbear- Riccardo s'lorandiperfectll'dealswith
ings,svhen 1)span isloaded itdet -lects the aboN'e problem s.Towers are ex- ln 1966-1968. Fritz Leonhardt pro-
downwards.butthistimetheadjacent tremely rigid.having the shape ofan posed :
1 proiect to cross the RiN'er
pylons turn alm ost freely because inverted V longitudinally.and can bal- Ganges at A llahabad in India. with
there isalm ostno tension N'
ariation in ance alone the effectsofasym m etrical aseriesofm ainspans.each 159 m Iong:
thecablestaq'sthatsuspend theloaded live loads.The installation ofinclined the bridge n'as about4 km long.The
span (Fig.9). Only the deck rigidity struts.which supportthedeck on each tow ers were again extrem ely rigid.
can balance the load effucts.Nvhile the side ofthe tower.isonlq'needed when with pylonshaving the shapeofinvert-

StructuralEngineering lnternational 1/2001


'
X
'
A4t)t) -'
x4t?t) spondingpiersand connected to then'
l.
I ' Adiacent cantilevers 5.
5erttjoined b)'
shortdrop-in spans.60 m long- to con-
j . ' I slitute aseriesofbal's.50()n-
llong (Fig.
*J t..
. .e*-
. .. J '
. 12).
..
lj ' ?
j 1 This project. directly inspired from
j'
. .: Nlorandi-sdesigns.wasaforerunnerof
1
the Rit-
)n-z
Antirion prttiect.and oneof
the first attcnnpts to develop heavy
prefabrication.Thistechnique laterre-
ceived many attractive applications.
'
' ) specialll'underthe influence ofBallast
l
' V Nedam . for example the Bahrein
1 ,
1 Coastu' ay.the GreatBeltqkstBridge.
'
i j the Second Severn Crossing in theL-K.
'
the Confederation Bridge in Canada
j w hich gives access to Prince Edward
lsland.and theO eresund Bridge,
The second project-directly inspired
fron'
lthe preN'ious one butwith n'luch
shorterspans-u'as deq'eloped in coop-
eration with Campenon Bernard 1* 01-
'-'j 1.'
;Q) ).4p ).
';;t) j.jt.l j the Ré' 1sland Bridge com petitio1) in
i !
'
j 1
: ' r '
1
.
j . 1 .
1986.Asin the previousproject-the
'
1 j bridge was made of complcte can-
. .
I
1
tilevers.totalll'in prestressed concrete.
MN A'
K '
ww. v.x'x' x.
-x M'
/ xNNw -' x '
r The slender deck u' as suspended and
, ./'
rigidly connected to 11rigid plr lon haq'
-
i
ing theshapeofani11y'erted V.longitu-
1
dinally and tranq'ersely. and was in-
stalled on thecorresponding piersbya
*
'
1 (
ed V longitudinalll'
.and piersdivided '
j
into tu'
o shafts.one below each ofthe 1
;' . lu
twopylonlegs(Fig.1l).Drop-inspans lh
k. X/ lX v.
zz.
were installed in each bay between .
y/,z.j
l.
p
:
'
. jj
1 '$
$. .. .
.. .
.X. x ' j' .è
.
,
$y x -.-- .
w
.-'NX-N.....
N'
V rv
,
,f p
cable-slayed cantilevers.to allow free ,. $u
ljN:. j.)' ,j
. 'j <.x.xx
,
$j .,
x. x..X
j x
A)Y
. rz/f.-- u.a u
length variations in the y'ery slender '
( .' 'uq..1j
r
< .c
c'' xc.s.s.sxc'.c .cxc.c..
.-kz
--: h$ j y tj
' i. -
..)
... -j.
=.:
,
-
.
. -
.'
..,y.'.sjZjx-.w
0K
,j '
x
ja
m-c
l
,
r:
u;
i
.q
mn
'. 3
k. ! ',
' .
deck. j mw' : rr
I
M ore recently.the contractor Grands I
' 1 !
Travatlx de s'
larseille (GTNII deN'
el- a'',#)t)()t, h()tltxJ -',3()t)d):i '
l1).()()t)
.5(
.
oped three projects.the firsttwo of
which had no m ore successthan those
ol
'Finsterwalderand Leonhardt. '

'
l'he first one. developed with SOF-
RESID and Jean-claudeFoucriat.u' as ''$h
partofa projectto crossthe English :
1k jyt
/r.l' l
h.
C hanndl.between England and France. j
, yyy-
,

atthe beginning ofthe l98Os.A Iong /-


?,
; j
1. .y.
y
y-

bridge was proposed on each side of j- yyyy


// . ;
the Channel.giN'ing access to an off-
shorestructure inw hich ahelicalram p
///I /: ( h
//
t' .. ) y
lead the traffic to an im m ersed tunnel yt/ s ! ïy %
,
r
that crossed the central part of the . '
h.
' j
Channel.Each ofthe two bridgeswas : k.
-Nu

m ade ofa series ofcable-stal'ed can- h


( h.:
J //
2
tilevers: the deck was a rectangular ! -.- (g-'
-
A.
()
(
)()
)(
)
r .1
1
orthotropicbox-girder.totallysuspend- . .
'- ;
ed b)'cable stays to ?1steelpylon haq' - h' -. ù ,'
!
ingthe shapeofaninverted V- longitu-
dinallyand tranN'ersally.Thecantilevers i
1 ! .
Nvere prefabricated and illstalled by a j.
:
1(p
.t)
(.
h
f
.Is
.
t
strong floating crane on top ofcorre-

StructuralEngineering lnternational 1/2001


tloating crane (Fig.2-?).Brith spansof
only 140m.thecantileversNverejoined
directlyN'iaanexpansionjointatmid-
span.The deck crosssection-designed
byJean szluller.had the shapeofathin
slab stiffened by m ultiplefloorbeams.
with the sidewalksata low er levelto
produce flexural inertia. The author
reused this principle with som e
am endm ents and im provem ents for
the Burgundy Bridge at Chalon-sur-
Saône.France.
G TNI cam e back to the ideas of R ic-
cardo s'
lorandiand to itsChannelpro-
%W =..
! ..%. '
iecttodeN'elop thepreliminarydesign
ofthe Rion-Antirion Bridge.Greece.
in thelate l980s.Jean PaulTeyssandier
and Yves M aury'steam designed four
large cable-stayed cantilevers. each
supported on a large off-shore caisson
w hich at the sam e tim e serves as a
foundation and a pier.D rop-in spans.
50 m long.are installed between can-
tileA'
er ends to com plete the spans
(Fig.144.Each cantileN'erismade ofa
pylon Nvith fourconN'erginglegs- with '
I *
'

the shape of an inverted V longitudi- 1


nally and transversally - and a com -
posite deck rigidll'connected to the I
;Z.
'
l
!
pylon and suspended to the pl' lon by
cable stays. Each of tlqe three main
1
i
I
/,
j-
spans.560 m long.isthtlsm ade oftwo j '.
.'
,
'
. I
m ain cantilevers. 255 m long.and a
drop-in span.Side spansare m ade of j
!
1
(
-
!j ! t
one cantilever.also 255 m long.and of ' I i
,
1
j
a drop-in span 50 m long to reach the 1 ï l
I
abutm ent(an approach viaduct). I 1
There is only one dil-
ference from
slorandi-s design: cantilevers (with t 1'
'
theirpylons)are notconnected rigidly j
1'
il '
l
i 1 ,

to the piers below. lnstead.to lin- lit


seism icforces.cantileversreston piers !
!;
'j
1
.j!
1I1
.
iE
! .
j
l
through a series of sliding bearings. l1'
j I'
I, 1 jj '
l-
with a systen'lof horizontaldanApers 1! 'j
to lim it longitudinal and transverse
.
'
t i .

displacelnents during earthquakes. It


w'illbeexplained below how thispre1i- i !
m inao'design has been inlprtlved to Fig./.?.
'Pl.
'
(ïj('
('
:fy/'f/pcSpislalltlb?-l
't/k
vt.
-11J'Ctllllp(?tïlll?BL>1'
tl(ll-dt???t/(-
,
-T'
.î.
/(l9:/5,
)
elim inate drawbacks produced b)'the
largenumberofjoints.bendingforces M
af,-1m <fn
'l1rl1 <fa()n4 76t-1n:
. 7()<m
induced by the concentration ofloads l
and cable staysatthe cantileverends. ! L
largem ovem entsin drop-in spans-and
com plicationsdue to theinstallation in /
arestricted areaofalargeseriesofbig . . . = = - - -
dam persthatneeded to bespecifically
designed and fabricated. >ta()m
-
n()
I .
DifferentTypesofSolutions
z, kx .
Solutions other than of the Nlorandi l;
- .- - = .- . -M .
= .- =- '> ' .. w ....
type can be developed. This seclion
review s the m ost prom ising of them .
even ifsom eappearineiegant.aesthet-

StructuralEngineering International 1/2001


icallyorfrolu astrtlcturalpointofN'ievv
(Fig.15'
j. Problel
-
ns related to deck
length variationslvillbe ignored here
in order to concentrate on the rttduc-
tion ofbending forcesinduced b)'liA'e
a)Internleuliattlsupp.ortclN'ctr),'sectpnd span loads.The design ofpylons- an axial
pylon or lateralpylons (one on each
side oftht)deck)- Nvillnotbe consid-
.A
ered here since this choice is m uch
m ore related to length variations than
to the balance ofliveloads.
/ / z
b)lIctatl-cc
'dblcs The firstso1ution consistsofintroduc-
ing an internèediate supporl in every
secondspan.x71isisthesolution classi-
cally adopted for susp.ension bridges
when two 01'three lllaiorsuspended
spans are to be erected.Some ttxCIm -
plesare fam ous.stlch asthe tw o O ak-
land Baysuspension bridges.U SA.the
twostlspensionbridgesontheKoiima-
Sakaide route ofthe Honshu-shiktlku
project.Japan (Fiz.l(
)).:
1nd the three
2
suspension bridges thatcross the Ku-
rtlshim a Straights on the O nom ichi-
uipC-ablevsl:1s.xcthm1ngfronlbk)thauijaccnTpvltpnx(t'xupporlthccQ'nlralpartoft?achqIa,an lm abariroute.Japan.
ltisnotalNval'spossible to installsuch
interm ediate supportsand the author
considers this solution the m ostinele-
gant for the design o1'bridges with
'
. . . . o . . <. .. . s. .. . xja.: .
m ultiple cable-stayed spans. Fortu-
'Z<--'...
C
I
,
C:I
..I
.-
vv. .m*,
4.
, .4
((.
w..
. -,
..
p'I-
a
.q
Hz
?#
:. .>
..:-
>.
..-
..
..r
'.s.
xc.u
.-
v.
%..
.. ..
..%
p.,
.' .....
-u..
.t .z
:a.
Eq
;.
z,
N
'..
a
r
z
.
zg
xy
r
?
.
'9
p.$ k
c.mt.
to
is.uslri.'
r.<.e.w. ';
.,,.....
,'.
-
a,
.o..'.y
s .hR
.
rx*w%.
.v..
..
...
.z
.a
'.
%.
:t
'.
t,
.z,
c:
.
.C
K.
rv.
y.
.D
N
.
w.
oax'.îS
: ;.
y.(
.-
a
#.
..
..
b-
-
e.z.
r.
;
1
.E
v
..
:..
4d,
8
'.
u.
t.
...
..
!
-.
, nately.nobodydared to do itso t' ar.
.. ... w . . .y
u The second solution is also inspired
.q. .. sw.
from suspension bridges. To preN' ent
'ho.- m:!'?,'.......,.
k ='
., .* /' pylons from bending towards the
::
'.. M.;:
....
& ''QeW'N .
. ' <'* x ' .
'I. *LQ' loaded spans.theirheadsare connect-
**x.r&..
.$
.-k.
tT..'a.r-z.
- !
' *
p
l
$.N.'.N
' .N. +'
%.
<-'%
*=.
'N'-
' ed by head cables.SeN'eralsuspension
.-.$'*.5.
p.
.xv.>.>.vh
.
..
.Jt%x.
YA.'
eI
. 'x
x>
<,..'' .
.;
*'..
.
c*
bridges u'ith m ultiple spans buill in
. . j
.. W...l
*A. Y-.N.<NN-'
ww-
. Franceduringthe l9th centurq'and the
.-
:xs-..
w.w..
h**.-'-
.A.
.'
+.'<A-Vk>
firsthalfofthe20th ce11tury have head
-.
..*>.v
*.. . WA. %'.k
...g.
<*+ cables.forexam ple the Sully-sur-l-tlire
<%
Bridge twhich collap.sed on a very
cold day in January l98j) and the
Chateauneuf-sur-l-oire. Langeais and
N'
Iasd-AgenaisBridges(Fig.l7).
Fig./6,
*1-11(
?ILçqillitl-saktlitiesldspf
. u
,
/lsl
'
ta/.
lbl
-idges.I&'
l
'f/Itlieil'f?'ft?l
-/?.
?ediatesl
.l
/;jptlr//pJ7#?Tt)H.'.4B..j) The so1ution l'nighlbe adopted lbrca- .

ble-stayed bridgesw ith Inultiplespans-


butitisprobablylesset -ficientthan 1
*()1-
suspension bridges since cable-stah' ed
bridges are l'nore rigid-and the llduli-
tionalrigidity provided by head cables
would be n' lore lil-
nited. ln addition-
the system looks lesseleganlthan for
suspension bridges.because ofthe in-
troduction of a new Iine that clashes
w ith the inclination ofcable stays but
is too sinailar. thus destroying the
structuralsimplicity ofclassicalcable-
stayed bridges and producing some
confusion.Only one projecltook in-
spiration from this concept. nan-lely
the one thatNvon the design conèpeti-
tion forthe Poole H arbour Bridge in
the LIK -whose erection has not yet
I-ïjr
.17:7-
/!t?xlI.
i
k
/?tz/l.
yj
'
fp?? been decided upon.

StructuralEngineering International 1/2001


A third solution isto introduce.in addi-
tion to the classicalcable staysdistrib-
uted along each span fron' lthe pj'
lon to
mid-span.long stabilisation cabîesthat
areanchoredon onesideat:1pylon head.
andontheothersideatanadîacentpy-
lon atthedecklevel.Thesestabilisation
cablesalso introduce a new line in the
structure. thus rupturing the harmo-
niousdistribution ofclassicalcable stays.
This solution was adopted by Jörg
Schlaich and R udolf Bergerm étnn for
% >fa6 >
the Ting Kau Bridge in Hong Kong 1
(Figs.2(
iland l9). A s this bridge has j !
onlythree towers.onll'thecentralone
had to bestabilised thisway.7-he com - m

posite twin decks rest on cantiles'ers


w'
lalerally extending from the axialtow-
ers through classicalbearings.so that
length variaîionscan develop freell'.
O nemightthink thatthissolution could '
.. . ' i' > a. J@.. .c . ' ... I.
JJ-
be im proved by distributing classical **'.x '
),j
.L ;.-f-z'., -. 'Q'z
'.
#zk3
.b.
. '
ï.
n.p
jtR
';
-.
'.
.
'.
......z -' l I œ-'9%
.' j
,
.
.<. ju<
lg.*rp
.-w. . c.1..<
cable stays from each pylon-1he last ) & '
onesbeing anchored beyond m id-span e
. . ' v e-
so lhatthe centralpartofthe span is . . .. .
. M '.
.
.,
.
. .. .
' *'
..
.
x., v . . ...c..e&
suspended from both pylons b'. 'cross- .,, '
%;; ..,..t -z
w -
ing cable stays.Th1: can be efficient /
'
. 7,/.?'
4g.
.
f.
t $
PCJ.;..,.;,.
r%'1,Y .
'
. '
...sy o
y
only ifthe deck isextrem ely rigid.be- zjzy
/
y(.yz./
Fj
../
lx
zj
. ..
. ..u
..
cause theu'eightofthiscentralsectic)n . ..xfz
.v
y ,,...'k..,.e#'e
.:..
#
ofthe span hasto be diN'ided between N
the cable stays that suspend it fron- l . ,
, * < 'A
both pylons. - '
##
z>&
- p . .'Y.
.
<
ln fact-this type of solution Nvas ini- .
. ..y
. > ,
. tt
. *F..
v
, y ., .
( .
tiated by Fritz Leonhardt and Jörg .
#'.
w ..A'
! xwysz
*
.ee < .'+
Schlaich in 197l when designing the . A
Patna bridge oA-er lhe riy'er Ganges
(Fig.2()).From thelessonsofthe Allah-
abad project-they decided to shorten
eN'ery second span by 2()' ?.
k
n (tlle span 4'..
q(#L L. 4).stIL L I,.yd1L
lengthsbecoming equalto L and 0.80
L alternativell').and to installcrossing
stabilization cables in these shorter
spans.Lengthvariationscould develop N
in expansion jointsin the main spans.
thatis every second span:butthe ex- Fiv.Jf-l.
'1'
11(:P(lê??t?I)ri(1zecF'
t
'
?5'
.
r
;/??cîlle,'
/1.
(,1-('
.
7t?F?ccs-f
. 'l971)
pansion jointsweredesignedtotrans-
m it bending m om ents in addîtion to
shearforces.asu'e shallsee latdr.'The
m ain spans were abotlt 2()() m etres
1ong : 1nd the totalbridge length w as
abotlt4 kilom etres.
Jtjrg Schlaich-w ho worked forthe A l-
lahabad and Patna proi
2
ects. reused
the concept for the Prlnce Edward
lsland 1ink com petition around l990.
u ith in addition an extensive use of
heavy prefabrication (Fig.2I). The
spans u'ere alternatiN'elq'220 and 180
m etres long.with stabilization cables
crossing in the shorter spans: short
drop-in spans. only 2(4 m etres long.
w ere installed in the longer spans to
free length variations.R-he proiectwas
to prefabricate 380 m etres long units.
StructuralEngineering International 1.
/2001 Rcporls 69
consistil'
lg ofa shortspan with its two
pylons. cable-stal's and stabilization
cablcs. and B' ith the t' wo long can-
tileverssvhich Nvillbe laterpartofthe
adiacentlongerspans:the closi1-1gbe-
tqveenadjacentunits-aftertheirinstal-
lation on piers.Nvas prodtlced by the
drop-in spans.
'
Fhis Iist n-
tay be conlpleted vvith tl1e
M acau Bridge.designed by Jose Luis
Cancio Nlartins.'wit11thecolIaboration
of Jörg Schlaich-Nvhich w' as inspired
by thisconcept.lthas tïvo n- lain spans
ll2 n'llong:btltNvith tu'o tou' ersand a
shortspan betNveen the tuo Iong ones-
itbehaveslike two independtlntcable-
stalved bridgesand thuscannotconsti-
ttlte a realbridge u' ith m ultiple cable-
stayed spans(Figs.22-24).ltwascom -
pletdd in 1994.

D istribution of R igidity
betw een the Structural
M em bers
-1-he best s0Iution-111-
1d the n'
lost ele-
gant.is to distribute rigidity betw'een
the differentstructurall' nenlbers (the
deck.piers.and pylons)inorderto bal-
ance bending effects produced by
asyn-ln-letric 1iN'e loads and tt) linlit
det-
lections.Frolu one extrenle to the
other.severalsolutionslnightbe ctlnn-
pared (Fig..
?-
$):
-
A deck with high enough rigiditl'tt)
resistbending I'noluents induced by
liN'
e Ioads.This solution is tlnly ap-
plicable for cable-stayed bridges
w ith snAallorl'nediulu spans.
- Pl'lons with high bending rigidit).
for exanzple haN' ing the shape t)l'
an inverted ' V longitudinally-with a
transfer 01 * bending forces i11 the
piers belou'to lil-nitthe rotation at
the deck level.LJ11der these condi-
tions it is possible to l'las'
e a N'ery
slenderand flexible deck.
- D istribute rigidity betq'ettn all the
structural n-lennbers-deck and tou' -
$1r:'i!.
A s stated above.the strtlcttlraldesign
n-lust alloNv for length varialiklns pl-tl-
duced in the deck by the installationt)f
continuity prestressing tenklons in the
spansaftertheyhavebeenclosed-con-
cretecreep and shrinkageuvhichdek'el-
op afterclosingthe spans.antlten- lper-
aturevariations.A llprclcticals()ltltions
u'ill therefore be analy'sed t'ne at tt
tirne.

'
The firstsolution is tt)giN'e the deck a
high flexuralinertia.large enklugh tk)

StructuralEngineering lnternational 1/21101


balance bending moments produced'
by asym metrical live loads (Fig.J6).
Pylonshaveaflexuralinertiasim ilarto
those ofclassicalcable-stayed bridges
and are rigidly connected to the deck.
The structure com posed ofthe pylons
and deck issupported b)'the piersbe- '
low through a single line ofbearings
on each pier.I1'loadsare rather lim it-
ed.thesebearingscan be classicalneo-
prene bearings. but m ore generally
they haN'e to be po1 bearings.O n the
two orthree centralsupports.depend- lahInlk?rn7edi2klktsolut1
't)ns.v'k1:.h rlgitl1t-
v'k.l1stribulcd bcluk)c:1p1..
Jrsd,2ckand '
ioAlttnx
ing on the flexibility of the piers and
foundations and on the span length.
they can be fixed bearings-butsliding
bearings are needed on the extrem e
piersto allow free length N'ariations.
This solution is adapted to axialand
lateralsuspension-but only for snnall
or m edium span lengths. The deck
flexuralrigidity alone cannotbaIance
asynAm etricalliN'eloadsinbridgeswith
ltlng orvery longspans. '

Three bridgeshave been builtaccord-


1 l,
' J
. j
) : i
ing to these principles.butdue to their 1
linaited span lengthsthey havereceived /
allnostno attention despite theirgreat
inAportance in term softechnicalevo- / zz- / - z-. - zx zz'
-mx .
,' .
,
<
1ution.
Thefirstone isthe Kwang Fu Bridge (
in Taiw an-designed bj T.Y.Lin and i
completed in 1978 ( g.P7)
.Fi .. I
t has l
;
threepylonsand two m ain spans134m
long.The deck ism ade ofprefabricat-
ed and prestressed concrete girders.
ithveryclassicalshapes.installedon
U' 1
.

temporarysupportsandjoinedbJ'lon-
gitudinalclosures.cross beam s and a r
t
.
!:
-#2.
-
a.r
I
:1svjyt
l-
n
''
.
cast-in-situ upper slab. Extrem e py-
lons. on the extrem e piers. are sta-
bi1ised by back staysin a classicalw ay.
eN'en though the side spansare slightly
Iongerthan usual.Onlythe centralpy-
lon issubiected to inAportantrotation-
all-
novem entsproduced by asq' lnal-
netri- 4(1n
...ilt)
(z- r..1 .. , ,. . - i
callive loads.w hich are contrclled by 9) ;a..
sii(,
1 ... ? lt) 1
the deck tlexural inertia. Pylens and 3
7
>. I .- 1 .
ù
.
'-
w. r: ! ; .
w.
cable stays are in verticalplanes.one '-
r 4
rx.
I '->
.4 >.
'
-
w i x x
on each side 01
'the structure (lateral . x' I
suspension).
The second example is the Colindres 1
Bridge-Spain.com pleted in l993.lt '

also hasthree pylonsand two m ain ca- ntà4Ij


-t9 o5() I>()() os(I ,!-=th
ble-stayed spans 125m long.A sin Tai- i i
f i 1
w an.the extrem e pylonsarestabilised ' I
by back staq's and only the centralp)'- 1 .
i .

lon is subjected to important moy'e- U- l


-. 1
Z l
i
n'lents produced by asynnm etricallive :
j i'
-
1 CI i
loads.which are controlled b),
'thc deck
tlexuralinertia.Pylonsand cabIe stays
'
! =ij(!I l
'
'
l
/
l
are axial. l 1(
,.
r
-() .
------..
---..-.----.
-z
j
I

StructuralEngineering lnternational 1/2001 Repklrts


7-he third and the m ost elegant ex- relatiN'ely small spans limit flexural -1-
0 improve the sl.stenn piers and p)'-
ample is the Arena Viaduct. Spain. problems-and the deck flexuralinertia lons m ust contribute to the global
designed by Jtlan Jose A renas.B, hich easily balancesthe effectsofasymmet- rigidity. The sin-
lplk)st solution is to
was completed in 1993 (Figs.2( $-.: /). ricalliq'e loads.The bridge elegance is
14, m aintain the preN'iotls design. i.e. a
Thisisa realcable-stayed bridge with increased by the S-shape of the road bridge with a relatively rigid deck and
multiple spans since ithas six pylons alignm ent. pylonsaboq'e the deck thatare rigidly
and five main spans l()j m long.The connected to it. H ou' ever. tht? deck
'
and pylons structurtl is supportttd on
l (Q (
:.
hyj (4:) o, the piersthrough tu'o linesofbearings
..F-''' v)
x-
j'
' (Ftf
?.31).Once luore these bearings
,' W
., '
yyk(!) y ' ? can be classicalneoprenebearings.bu(
,.
,
/'
.
o/ o .
, î.' 1 ''' n'lore generally and specially forlarge
,
I
)
t
$,
'
. f ' loads.they n'lightbepotbearings.fixed
' '
on the tu.0 01-three centralpiers anul
. . $8
slidingon theotherones.
/ %k. X.
H oNvever.lhissolution doesntltstllq' e
colupletely the problem oflength N' ari-
ations.A't first. if y'
ertical ltlads are
very high (e.g.forlong-span bridges).
high horizontalforcesdeN'elop on pier
heatls due to friction on the sliding
bearings (it is necessary to cotlnt at
least2 or 36-
)t' for the friction coeffi-
cient).lfthe piersare tallorN'tlrl'tall.
a'.* $:
thiscan be prohibitiq'e.xAnotherprob-
.. ..i')t-,
l
; I :
len'l.which can be m astcred m oreeasi-
: '
1,:.is the ilmportance of the relative
r I
m oNr
em ents of the deck (and p)'Ions)
on thepiers.ltisnecessary to consider
r these movements in the design since
'
j the reactionsfron- lthi?sliding bearings
- fixed on the piers- nnove by tcnsof
centim etresbelow the deck.The cross

y
beam sin the deck above the stlpports
and theirreinforcem enthaN'eto be de-
signed accordingly.ln addition-though
thisisoflil- nited practicalim portance.
itis inzpossible to predictthe relative
luox'em ents of the deck on t1e stlp-
ports:because they arecontrolled only
by friction k?n the sliding bearingsthey

T cannotbe evaluated directlysillceslid-


ing n- loNv
en-
lents are non-linear and
non-reversible.These n- loN'en-
lents can
be onlyevaluated afterselecting a val-
tle for the friction coefficient and a
conzplete time history'of loads and

Reptlrts StructuralEngineeringInternationul 1/2001


length variations (temperature-con- controls flexural rotations very well p'oints to allow forlongitudinalmove-
cretecreep and shrinkage-etc.).lfthe and producesan alm ostrigid connec- m entsin som espans.
relative m ovem ents can be very large tion ofdeck to piers.H owever.italso The solution was adopted in another
(about0.5 m fora very long bridge). m akes the longitudinalm ovem ents of projectin Denmark.namely in 1972
greatcare m ust also be gisven to the the deck alm ost free ifthe shafts are
forcrossing the Sam sobelt.butNvith a
shapesofpiers and deck to avoid evi- individually flexible. continuousdeck.The bridgehad three
dencing the off-centering.These rea-
pylons and four spans 264.624.624
sonsexplain why theauthordisregard- It is thus possible to design cable-
and 264 m long.Extreme pylonswere
ed this solution when he designed the stayed bridges with m ultiple spans stabiëised by classical back stays and
M illau V iaduct.France.which ispre- having a relatively flexibledeck.rigid-
onll'thecentralpylon had to behighly
sented below.even though two linesof lyconnected to rigid pylonsabove and rigid.Piersm ade oftwin flexibleshafts
sliding bearingson rather strong piers topiersm ade oftwo paralleland tlexi- m ade freethebridgelength varialions.
could have appeared asa possible de- bleshaftsbelow (Fi g.- i
U).The bridge the im portance ofw hich w asrelatively
sign. length is.however.lim ited bylongitudi-
m odestat the extrem e pylons with a
nalmovementswhich can be accepted
dilatation length ofonly624m .U nfor-
A n extrem ely interesting alternative by the extreme piers.asthey are gen-
existsto the solution with two lines of erally shorterand thusm orerigid. tunately.thisprojecthadnomoresuc-
cessthan the othertw o evoked above.
bearingson the piers:piers are m ade
of two parallel shafts at a relatively This solution w as identified N'ery In a completely different approach.
smalldistance from each other longi- quicklyandFi nstenvalder'sprojectfor the deck can be cutw ith an expansion
tudinally.Asshown by JacquesM athi- the G reatBeltBridge had piersm ade joint.buttheclassicalarticulationin a
vatatthe beginning ofthe 1960s.for oftwin shaftswith a rigid pylon above- m id-span cross section. which pro-
the erection ofthe Choisy-le-Roiand as di d Leonhardt's project for the ducesexcesst-lexibilityin the structure.
Courbevoie Bridges in France.a pier G angesBridge (see above).H owever. is rejected.The principle consists of
m ade of tq'o parallel slender shafts in both cases there were expansion producing a continuityofbending m o-
mentsthrough the iointorjoints- in
the mid-span cross section of one or
severalspans- by introducing a draw-
erbeam inside the deck qlï
-ig..
3.
?).This
'r
'
Vi '
drawerbeam-which has to be in steel
to lim itits size and weight.issupport-
L-
'
!
-'' ed up and down on tNvo linesofbear-
ingsin each ofthetw'o cantileversthat
/ / Z X // / A itjoins.Itisfixedinoneofthetwocan-
' tilevers. and slides inside the other
j
rn
' one.Jean Nluller developed a system
.
I k/ /
? ofthistypefortheR ogervilleV iaduct.
i
- r ,
'
! //$k ! France.a classicalbox-girderbridge in
1
, '
,
:
$ . jt
./1
jy
yk
j .
,
'
r ( l
j / 1
' 'h I prestressed concrete.w here the pier
( ' . ' j .
1
( I . I
1 zk
'r î i
$
' architecture - structurally illlogical-
! r
î i ,
j.
1
. j 1
' . à.t' m ade such a design necessary.A sim i-
.

1 I 1 J1 lar concept was adopted by T.Y.Lin


, i1I l t j 1
. lnternational for the new Benicia-
i
X
s
:.
1V
.4N
..>.
=''.
xmkw-.v
<.
-$+s. x
'ks
;.=.5l Q%@' w%w
.*
-xw<<wl v -' .'.'>f.wr=:'>x .<&x
.'=
w<W <e1.
.v ,
'*.zw
vur
.arcw
x=v.
x - ..
g la.m.q'w..- lqx , xr.w Nlartinez Bridgesnow underconstruc-
.A&'>* w>
4M
>v .wrU%>.'
.
WAx-
.x -'. *=*
'xs'amv:x#h.':xa
.' .*
- . ..Xe .>
, . ... <=
. 'v.
==p' - 'a
. '=?
tion in the U SA ,w hich are also classi-
cal box-girder bridges in prestressed
concrete m ade from precastsegm ents.
Such a solution is only feasible when
the deck.preferably a box-girder.has
dim ensionslarge enough to house the
drawer beam .and has been designed
from the start considering the corre-
V Z . V ZV V Z# spondingconstraints.
'
'
1 j
l ! A nothersolution isto com pletely sep-
( 1 T
1 I ' j l arate the deck from the pylons.The
: i pylons are then directly extending
ï
d
1 j
I
piers to constitute towers qFig.J.
/).
:
j ' This solution is extrem ely practical
)
w hen the tow ers are m ade of tw o
1
eolum ns-one on each side ofthe deck.
l !. . 1. ltcan also be foreseen with axialpy-
om'
..
e' v.'e*=
Yv$< 'v '
aze ..
- --v..
a. .q.
.u
y o-a
%
. . eo
u.
x.*..
. xw
y=x*x.
ba.-=raqH
s.
'-
..
2F- -'
>.'
-l
y=.v
+rGv'
. +zYr
-zwel
V'
wx-
..r
kp
m-4
-....
..o
e<
a+
.-..-.-.
.1o.
tP-=d
y
>;cz.. :z
%zLr
.z
=zJu
>'
a'
'ok'sw
w! ,=s(C%,'.s
A.'-
N+
''
=
%A'
@M. z
.
u. .4.
98%
tm=.
<X.
k '%>
Mq-'x
tv 'l
'kx''.
i'F
e'cw.
*. rr
vsx=L@:
'k lons that pass through the deck.in a
' j ;
hole wideenough to allow forpossible
relative m ovem ents.The deck is sup-
ported on the tovversthrough bearings

StructuralEngineering lnternational U 2001 Reporls 73


-
fixed orsliding.dependingon thepo- ln such designs.length variations are -
are com pletely separated. On the
sition ofthe tower in the bridge - in- notlimited b)'the towr
ers'rigiditl'
.and other hand.the load transferred l'
ronz
stalled on acrossbeam when thetovver the two main problems - bending the deck to the pier (the partofthe
isaportalframe.ordirectly on thepier forcesproduced by asym m etrical live tower belosv)is rather low since itis
if the axial pylon extending the pier loadsand length variationsin the deck linnited to self-qveightsand loadscorre-
passesthrough the deck. sponding only to the partofthe deck
close to the tosN' er.The Iargestpartof
the loads is transferred to the tower
heads by cable stays and frolu there
passes directly through the tou' ers to
the foundations withoutanq'interfer-
. m . z A Z ence with the deck.Friction on sliding
bearings is thtls lil- nited and cltnnot
generate high butnding forces in tou'-
ers.
Forthisreason thisisan extren- lely ef-
ficientsolution which wasadopted for
the N'lezcala Bridge in Nlexico.com -
pleted in 1993.with three toNvers and
tB't) m ain spans 3l2 m long (Figs.
J.5--:
?7).D ue to site ctlnditions.the ex-
trem e pylons.which are stabilised by
classicalback stays.are shorter than
1he ce111ral o11e.The con-1posite deck
passes freely through the two legs of
each towerand issinlply supported on
the pier.Nvhich constitutes the Iower
partofeach tow er.This and the Ting
Kau Bridge are the (lnly exaluples to
I date ofcable-stayed bridgesw ith m ul-
tiplelongspans.
The single drawback ofthissoltltion is
its lack of elegance. with the two
co1um ns of the tour ers fral-
ning the
structure and giN'ing ita ratherstockl'
appearance.The uppercrossbeam be-
tw'een the tw o colum ns does notim -
.k
.*vs
$%t
Obw
G'
Q:
:::
a......
.Q.
!Z ''.A'
R
'g.=.H..
'*M' ''.
T.c'> w
tc
=f.1+.
zq cy- p<.'
$* ?-
/.*'-
''
' prove it.nor the w ide pier btllou'the
;
deckwith itslateralextensionsto driN' e
the colum nson each side ofthe struc-
ture.This is the reason forthe choice
of axialpl'lons and cable stays Nvhen
the author designed the Nlillau
V iaduct.

v- : The Ting K au Bridgt r follows thesu


;. i lines Nvith free lengt1) N'ariations and
... //,.$$x : ,, ïs--s-s
?li
h X---x .
t :
j
.. . , y -. j axialpylonsfor greaterelegance.' The
.I
i /!
. : k :
il ''' e'
'
t4
j--w'xwxN .
'
-Ag4
.'j)() h?
j ..'
deckism ade oftwo parallelcom posile
StructklreS. eaCh con3priSing tNN'(.
3 Stee1
--x
-7
l'
--x 1-j,!
, -- .uk'
I &
j -:1i.'av
l
i .- I-girdersand areinforced concrete top
! '. E7t
lttlI'e kltlll15s't,terle)-.qjk-'1),
x'
w<r. < slab. connected by a series 01-steel
cross beanns thatare extended in the
'
!9.l()f)
.
twin decks as floor bean- ls.The axiaI
:'-
, l 1 pylons are installed in the open space
. !
e-
tpJ
! I i between the tw in decks. suspending
x. I :
1 .
then'l through four planes of cable
'
j -.. ( stays.The twin decks are sinRply sup-
.
---- c7 l
$ '
1 l 1
: j : $ k l -' ported on transverse corbelsextending
n. I xm ; ë F ,
, = 1
xc ,
.
:
1
,
!
t
;
1
:
,
.' from the pylons.so that length N'aria-
/ ; .
+
.
1 .L è !
t
tionsare free-only linRited by friction-
r i and loads transferred to the corbels
: ow441 -.
.
.
1j..
j.l()() l are ratherlow' .N evertheless.thepylon
strength is increased by a transq'erse
cable-stayingsl,stem .
Reports StructuralEngineering International 1/2001
l'
-
igs..
?6-.
?7.Tit't
'
pJ.
'
ïf
?3.
$l
.
'
kof#?t?fklezcalaJ.
l?Yt-
/'
(
j'
t:
-fp/?t'nl
sFrelssilîeîd
???J.4.t-
7?
t?d/li
ll/

'
This lastfamily of solutions leads di- inspirationNvastakenfrom thisproject A san approxim ation-and w ith an un-
rectlyto a new conceptoftotalsuspen- for the developm entofthe design of known hyperstatic effectgiven by the
sion.This concept w as developed by the Rion-A ntirion Bridge. value of the norm alforce in the m id-
Fritz Leonhardt for classical cable- span cross-section. the compressive
stayed bridgeswith the erection ofthe force in thedeck duetothisload varies
Pasco-Kennewick Bridge in the USA Sim plilled Evaluation ofForces from zero.in the m id-span cross sec-
in 1978-and reproduced in 1986 b)'Pe- tion.to qLll 'bàl'
l near to the pylon or
ter Taylorfor the A lex Frazer Bridge At a meeting in Tokyo.Jean Schm itt tow er.
nearVancouver-Canada.' T'
he concept gave usa very simple way to evaluate
adaptsperfectly to cable-stayed bridges som e forces in bridges with m ultiple W hen analysing the effect ofperma-
with multiple spans in that it permits spans.Considering thatthe deck isso nent loads in a cable-stayed bridge
completefreelongitudinalmovements flexible.we can neglect bending mo- with m ultiple spans.due to sym m etry.
ofthe deck - lim ited onlyby thestress m ents in itw hen analysing the global pylonsortowersareonlysubiected to
variations in cable stays produced by equilibrium ofloads.NVe could extend com pressive forces.'T' he norm alforce
them ovements-withoutany directin- hisapproach.supposingthatlongitudi- in pylons or towers due to the deck
terference with the rigidity of towers. na1deformationsin the deck are very Nveightis equalto PL w here p is the
Length variationsdue to tem perature. sm all. when com pared to horizontal deck linear wei ght (including equip-
concrete creep and shrinkage are thus dellections produced by bending m o- mentl:horizontalforces due to the
com pletely free (Fig.-$W). m entsin pylonsortow ers.even ifthey w eightofcantileverson both sidesof
are very rigid. each pylon or tower are balanced.
Thisefficientsolution wasproposed by N orm alforces in the deck vary in a
the contractorBouyguesfortheR é1s- For simplicity- we suppose that the firstapproxim ation from zero atm id-
land Bridge competition in 1986. deck ishorizontal.butitwould be easy spansto /pf.V8/?at the pylons or tow -
Pierre Richard had theideaofforming to introduce a correction w hen ithas ers:only construction effects (includ-
the bridge.about2.8 km long,from a some inclination. ingprestressingforcesifany).temper-
series of cable-stayed spans 210 m ature variations. concrete creep and
A uniform load on a cantileverarm is shrinkage can alterthisdistribution of
long.The deck w as continuous from directly balanced by the verticalcom - com pressive forces.depending on the
one end to theother.totallysuspended
from the towersthrough which itpass- pon entofcabletensionsqFig.404.con-
sidering the global load on the can-
type ofconnectionsbetween deck and
es (Fig.J9).Unfortunately.just after tilever arm .itis balanced by the ten- pylonsortowers.
thesuccessfulerection ofthe Bubiyan sion in the w'average'-cable-stay with: The situation is com pletely different
Bridge in Kuwaitand when the erec- for live loads.The case of a uniform
tion of the Syllans and Glacières T = qL load on a com plete span. on a deck
Viaductswas beginning in the French 2sina which is totally suspended from the
A lps.Pierre Richard designed a three- w here L is the span length and c the towers.orrestingon a1ltow ersthrough
dim ensionaltruss forthe deck in pre- inclination ofthe''average-'cable-stay. neoprene orsliding bearings is show n
stressed concrete. the high cost of The pylon or tow er thus receives a on Fig.41.Towerson b0th sidesofthe
which killed thesolution. horizontalforcegiven by: loaded span receive a horizontalforce
Bending forcesproduced by asymmet- ql- equaltoqLï;/8h.
.sincenoother(impor-
ricalliveloadsw erebalanced easilyby F = F coscz= tant)force can come from the deck
the high flexuralinertia ofthe deck so lttqa due to the type of connection.these
that the towers were rather slender. towersreceiveabendingm om entw hich
M ore rigid towers. not very m uch varieslinearly and isequalto:
m ore expensive.could have allow ed . 2
ik
(jL (H + /?)
for a slenderdeck w ith a m uch low er l(t'
?)= 8/ .
costthan the proposed one.the total ?
suspension m aking the length varia- atthetosverbasis.where H isthe tow -
tionscom pletely free.A sshown below. erheightbelow the deck.

structtlralEngineering International 1.
/2001
A nd since the deck can receive no
(inlportant)horizontall
brce frolu the
toïs'ers.through the bearings.the nor-
m alforce produced by live loadsatthe
' toNversisequaltt)zero:an hyperstatic
t
1 effectdex' elopsi1
-
1the loaded span.and
'
j l
1 the norlnal force is a tension in the
I lnid-span cross-section of the loaded
span.give1' 1by:
3
?'hh; j!
l- é
f'
t.
!1--
1 / z. d
r' ......'.......s

I
l 1 .2 8/?
1 j' leading to the tlffectiN'e distribution of
.Y.a.-4t-v.
Y sqav.w ! k..:n.x.k ... .. j l
* *.w ... *=.Aa.nF.
N*
...' -5a
s .Qr.
0=''r*1
..
. '.s..;w>Qm-:z*
a N. 1-r
-'. +g+os..Cm..
>,. .
*5#
>
*
w
rv
-'
eu*
.g
ns
'x.
>+
-'a.
.<
Y-
A
.sc
î
s
%s
*
. g)p.J
.t.*
.j.$j.'
- ...ï...,
.-- i-. tl. norm all'
k'-5tzrlxtjj orcesin thedeck produced by
, 1:
'
. 1 !j.
4.
c.-
oa. these liN'
e loads.
lfthe deck isrigidly connected to clas-
sicaltoNvers.or connected to classical
tow ersby fixed bearings.lhe situation
is conApletely different:a fram e effect
developsintheIoaded span and thead-
iacenttowers (Fig.42).Supposing fol'
sil-
nplicity that they have the sanle
height below the deck.the horiztlnlal
displaceluentatthe deck leNveliseqtlal
to zero for sl'nlluetry and due to the
fact that the longitudinalJeforl- natitln
inthedeck isN' tll-y'slnallasconèparedto
the horizontaldisplacenèentsoftowers.
The horizontalforce in thc touqt zr pro-
duced by the load-t/ '.
J-2,
'
8/?.is --balan-
.

ced--b)'a reaction A atthe deck leve1


with:

. -r
N xx .. .j 1 ..
..Nxwx N$
x . g/ .
y
y /z -.-'''.
-
e- I
.
N ..h N, ,z '$4.. ! .e-'''-
...e-'
Sc.
.x
'N-.x >X ,, .y ,' a
Nxxks. ;,z
. ' ''.I
' .
/x w
, -.
.ee 'I
.
'u=.
' .
j (5)
j wj
.' ..-
' . 1
*
d
' '.
.
. ' !
î
(
,
J 4.l()
(i j
'
The distribution of nornlal forces i1-1
' the loaded span directly derives fron-t
?
/
'
,--/h
!
-----
t
'
-
:
' . !
-.
--
-''
.-
- f
q
i
j
.

i
-
' ''
'.-
..
...
j7
thisresult.
ltisclea1'thatthedistribtltion ofbe1,
1tl-
??---
3
-
/
,
r
,

.
-,
y p
'
;
j
,
y
, ,
ing forcesin the tosversisnluch nlore
favourablethanw ith atotally stlspend-
/
j
.
'
-
; .
,
) t j
; !
,

?,
y
eddeck(ora deckon neop1-e1'
ing bearings)'
1eorsIid-
.at the tower biasis.the
'i
k. .
bending n'
lonnentisgiq'
en by:
xXx'x t,.'v.
'
y Sj ' -
...
..
j .
4, l
j
!:...!.:.kt ,-.I:.!,tkt .
l
I (6)
-
i
-
)-
1
.
-
l j7
'.1 j1j.
But.practically-u'henthe deck isrigid-
1).
'connected to toxverstheirlow'erpart
is organised so asto free length varia-
tionsin the deck.for exam ple by div-
StructuralEngineering International 1/2001
L11--
I ''n l
1 tvj( j
1
1 X '----A ,
.

'
t
r g;
r; ) .
r
l t
.
/y s.
Nx. *.+* =N v t 99, : , vj: j
srq l ! 4 <
&N x. V1 .
'.
.
t q' i
- -- - -- - - >-. .y- -- j
t 1 qj.-
.k b /-/:
-
1 &'?'E f/:. )./)('M
// .
- .- (3 -J..
-2
cn
L 1

Fig.41:I-FT' :Ioadstp/?a t.
-t?F??/?Jt?rt
7spall.It'JJ'
/7a Jt
?c-/k
'colltpleselj'-
çI/.
!'-
/7el1é/(,(1ri)/(?It'
Q,l-
s
'
t t?r/T
-'
t/L-i :
1 w, ' .
. - z.q/? j
; w (* . , ,? i
. I ! t,/--
l
I l ! 9
j: /
? ' k Jt I
' No /? R l t
?/-' , ( ,
j yk t uvj: (
fp..
-i I += = -lR
-.
&?
A== C
IL.. * ' --- ----- @
. -. yi. -
*=.i=-- s;: ('D
l ; y
( s' y j
I I
H I
B t//.S . z : de
zfz
z' /'ze
. z'.' l#)
qL'
(A - ? R
N'( 1k
l/3
a') (.>

iding them into two parallel shafts


(Fig.4J):below the deck.towers do
notresistto horizontalforcesbutonly
to bending mom ents:we can then as-
sume in a simplified approach thatthe
shearforce is equalto zero in tow ers The tower supporting the loaded cantilever arm receives the deck'
below the deck.Thuswe have: a horizontalforceRiasindicated in formulae 10and 11-below.
2
a Ii. 11
R= ,71 dx adx
8/?
The bending momentisconstantin the
If=
sy j(/?+H,--
/?
t.
'1(/'
f,-a')sy-R,J(H,-A'
)EI <)
lowerpartofthetowers.and thedistri-
bution in the deck of normal forces
produced by live loads îs the sam e as
forperm anentloads.
The analysis is m ore sophisticated
when only half a span is loaded.The
sim pler case this tim e corresponds to
a deck rigidly connected to classical
towers (atleastpartlyl-or connected
to classicaltowersthrotlgh fixed bear-
ings (Fig.444.Neglecting longitudinal
deform ations in the deck. it m oves
horizontallyasaw holeon adistanceu.
A and finall),.
ll towers rigidly connected to the
deck (orthroughfixedbearings)-ex-
cept the one w hich suspends the
loaded cantilever arm - receive from
the deck a horizontalforce.R:.given
by:

structuralEngineering lnternational 112001 Reports 77


From thisresultand formulae 7 and 9 zontally as a whole on the distance l/. This evaluation is an underestinnation
itiseasy to evaluate bending forcesin ltisalso assum ed thatpylonsare rigid due to the effective horizontaldelec-
alltowers.and also the distribution of and have horizontaldeflectionswhich tionsot'pylons.and to tht tverticalde-
norm alforcesin thcdeck.compressive aresm allascompared to !f.Thelength t'
lections of the deck w hich lilnit the
on one side of the loaded cantileRrer variation in a typical cable-stay in a tension increasein internlediate cable-
(on the displacem entside).and tensile cantileverdetlectingdoNvnwardsisthus stays.B tltitcan beconsideredasat' irst
on the otherside. given by : approxiluation.Fr0n' 1it-bending forces
can beeN'aluated in pylons.1'1'01,
1,
1ahor-
lf tow ers are verl'flexible below the (I+dl)2=(.v+!/)'+/?.' izontal force at the cable anchorage
deck as regardshorizontalforces.the
shearforce isequalto zero in thetow- u'here xiisthe horizontaldistance be- levelgiven by:
er supporting the loaded cantilever. tween the anchorage in the deck and
and again: the tow er in u'hich the cable is an- H = - 2 ,, s$ '. cos?a !/
chored.And t1 3tls: l '
R ql-2 ?=i l 1
= #/= ltcosa
8/? in typicalpylons-and b)':
where (.x,isthe cable inclination to the
isgiN' en as a firstapproxinAation:it is /-/= (IL2- , ys
notpossible to evaluate the horizontal horizontal. The tension variation is
disp.lacenAentwhen considering tou'
very flexible.
ers
then giq' en b),':
. . 8/??.)
f
-.
)(/j,cosza,//(l7)
in the pylonw hich supportstheI02ded
The analysis is m uch m ore difficult cantileN' er arnn.
when the deck is totally suspended
from all towers-or supported on a1l The deck horizontaldisplacenRentcan
tou'el's through neopren: or sliding be overestin- lated by considering that
int er nèediat e cable-stays receiN' e ()1'
1l)'
bearings (Fig.45). M' hen loading a clR = E.î cos-
' -
a lf N'er y sl
-nall t
ens itl
n N'ar iations due lo
cantilex'erarm -the tension increase in / î
the cable-staq' s which support it pro- the dec k 1
-lexi bility - and related vert i-
duces a horizontalforce in the deck. 'The san a e resul t is obta ined in a can- c al detl ections - and b3 ' tlsing f
or nlu l
ae
:./1./8/?-w hich inducesa horizontaldis- tilever arn-
:. ldeflecling tlpwards.ifthe l4.l5-16Clnd 17.svhereonlythelongel -
ptacem entofthe deck.dd.The vertical hor i
zont al act ion is counted in the c abl e-stays are cons ider ed.
delbrm ation of the deck corresponds sam e direction.N' Vith 11cable-stays in
each cantileN'er. and N towers. the Finalll' .Jean Schluittnoted thatu.l -
len
to a downwardsdeflection in halfcan-
tileverarm s.including the loaded one. gl obal r eact ion is given by : a cant i
lever is l oaded in eveo' span -
and upw ardsdet -lectionsin otherones: R al u-ays on the s ame side- hor i
zf-l
nt al
,? sy g forcesaretobebalanced in each tou' er
the vertiealdetlection is very sm allat
m id spans.For thisreason the longer
=
zj
s'J)(jjcosa.// ( Fi g. 464 .Onl ) 't he shear for
deck atm id-span can then balance the
ce in the
cableswork asbackstays.and they are
the mostefficientto limitthe horizon- Si nce t his react ion has to bal ance the loads-with:
taldeck m ovem ent. hor i
zont al effect o1 -li
ve loads .itcan be
concluded that:
2
For this reason.eq' en iffar from cor- (1kL
rect.it can be assunAed thatthe deck 11 =
l6 r/ i f yy 2
relnains horizontal and n- loy'es hori- kN ? - cos a
/ ?

(Fè
i<
i2/'2
'
* ==+ S e += $
i k
t ;
k I
1 . 1 y
'
'
i l j h'i .
:
' 1
i I
. !'
'
.;
)
2
- - 1- - - - - -- - - -
1=
- 4-- -- ; --
'
i
'
.
î = 1=
. s '
l '
. î . I
. (j t/ t/ t/
: '
in =
'
1 k Zz'
z /
'z
..
z/ //
/ :/
zzz Zz
(E) /
/ *2 yr2:
j ! i)i. :
'
i m $
l =+ >,/? ..
>; '*.
R ' X i ' !j.r 1Ik
=5# k. .z y..+ ,i..--
tt V
i I//. 11'
i // 'L.L? ' / a
< '
' lï-
ig.46:LJ ''
k'(.
z10(1(1.
b'tJ/2t?t' tg/lr/'/tJ'k't.
p/' tl!'ll2 111
Fig.. .1-.i!
/J. .
'f
?1()(1(1.
%(?/?a ('f//gfilej
..
er/? '
,.F?2.bb:il
'
lï ctlc/?A'ptfll-Iî'J'r/2/?(I(?c.k (-t')/?2/?/tzf(,lj'. $;df.
b'
.I
)(?11(1.
-
(1deck crp/??
p/t,
zftl
z/ï'sltspellded Jtp/ '
6,1$'el'.
b. L
'(IJ'
t.
lt'
()3b'c1.b'

StructuralEngineering Internatiflnal 1/26401


MajorProjects
Tsvo projects developed in thtlearly
1990sproduced importantprogress in
thedesign ofcable-staqr ed bridgeswith
m ultiple spans. nam ely the Nlillau
Viaduct-which crosses the deep Tarn
Vallel'.and a bridge over the lake in
Geneva.Switzerland.
The author prepared the conceptual
design of the N' lillau N'iaduct in
1990-1991.and developed the prelinai-
nary design in 1992-1993 u' hile still
working atSETR. A .H ou' e:'er.despite
tht)supportofthe localatlthorities.the
ProjecthasprogressedN'erysl()n'll'due
to man),diffictlltiesinduced bq itsil-
n-
portance and cost-and b)'som e oppo-
sition.Jean François K lein and Pierre
51oia took son' le inspiralion grom it
to design a cable-stayed bridge across
LakeGeneva.-l-heirproiectwasaward-
ed thedesigncom petition orgallised to
by-passthe city ofGenes'aon theeast-
ern side.w ith acrossing ofthe lakevia
a tunnelorbridge.Theq'u' erein charge
of deN'eloping a detailed design in
1993-1994.This excellent design in-
spired the Ylillau proiect from l994
on.so thatboth projectshelped each
other.

Lake GepdrtsBridge
The Geneva Lake Bridge is nlade ()f decide upon construction. and the H owever-thiswasagainstthe author's
four pylons and three m ain cable-
G eneva poptllatitln opposed in 1997 ideas.asheprefered having no tensile
stayed spans350 m long (Fig.47).The any proiectacrossthe 1ake. force in the m ain m em bersunderper-
alignn-lentisslightlyctlrved to increase m anentand frequentloads.Theauthor
thebridgeeleganceand m ore speciall3, thus preferred to divide the extrem e
to im prove the view thatuserswould slillau Viaduct piersinto twin parallelshafts in order
have of the structure Nvhen passing
along it. Despite this curvature. the Th eNlilIauprojectiseq'en l uoreambi- to produce high rigidity for bending
tious.Itisabklut2.5 kn4long.Nvith ' the m om ents and large flexibility with re-
pylons and cable staq' s ltre axial.The road passing 27()nl abox' e the Riq'er gard to longitudinalm ovem ents.Em -
deck isextrem ely u' ide.33.46 m -w ith Tarn.lt1' 1asses'en pl'lonsand six l '
nain m anuelBouchon increased this longi-
an extrem elj'elegant cross section:a cable-stayed spans 342 n: long.Nvith tudilèaltlexibility by installing oneline
trapezoidalthree-cellbox-girder-alnnost two piersn-lore than .' 23()n-ltall.the py- offixed bearingson top ofeach ofthe
triangular.extended on each side by lonsrising 9()Im above the deck.H on'- tw in shafts.which w asbetterthan pro-
w ideoverhanging slab elen-lents.Rigid- eNvel-.as stated aboN'e.developing the ducing a rigid connection as on the
ity is distributed efficiently between
thepiers.thepl'lons-and the relatiN' ely pl-
roject totlk a ' h,
rel-
l
y long tiJme and othersupports.
nanyproblelushad tobesolved.
slenderdeck. Not conA'inced by this proiect and
The prel il
ninary pr oj
e c
t wa s estab- seeking som e com petition betn'een
Longitudinal deforn- 1ations produced
by concretecreep and shrinkage.sol- ne I ished by SETRZ A.n'ith an- laiorcon- differentideas.the Road D irector tl1'-
cern about longitudinaldeform ations ganised two com petitions. The first
Prestressing effects. and tenxperature (Fig../*).711e idea ofan interm ediate one took place in 1993-1994 to select
variations are lim ited b)'the relatively expansion
shortdistance betw een thecentralsec- ing the r.ecom iointwaselim inated follow- new ideasand concepts.D esign offices
m endations of R ené and architects were consulted sepa-
tion oflhe bridge and the extrt an'le p)'- Xvalther.who was m em ber ofa panel rately. but very few neNN' soltltions
lons(475 n' l).'
Theq'are allowed for bq'
1e relatiN'eI3,high flexibility ()1-piers of
t1-
.
expertsinchargeoftheprojecteval- em erged.A second com petition u'as
uation.as he had been for the N or- organised in 1995-1996 betueen fiN'e
svith regard to longitudinaln- loN'el-nents m andie Bridge. René Nvalther would team sofclesign offices and architects.
and by the foundation conditions in haN'e accepted som e cracksin the ex- each team being in charge ofdeN'elop-
the RhonealluN' ialdeposits. trem e piers.resulting from length 5,' :1- ing a proi:
ectcorresponding to one of
U nfortunatelyforthe bridge engineer- riations. which u'ould l' 1aN'
e relaxed the t' iN'
e taluilies of solutions selected
ingcom m unity.avotewasnecessaryto bending forcds in these m em bers. afterthefirstconsultation.Thecolupe-
StructuralEngineering International 1/2001
tition u'asthusmorebetwzen proiects ofSogelerg-Europe EtudesGecti.Serf îectbctB'
een 1996and l998-initiallyto
than between teams. The iury in and the architect Sir N'orman Foster prepare a callfor bids betu'een con-
charge of'the choice in Jtlly l996 se- (Fig.- 4#).vl-his team which the author tractors.HoNveNver.itNvas recently'de-
lected thesoltltion '
with mtlltiple cable- i
oined afterleaN' ing SE-
I'RA wasthen cided to award a concessitln for the
staqr
ed spansdevelopedbl. 'aleam made in charge ofdeveloping adetailed pro- erection of the section of the nlotor-
B'
ay'ofB'hich theviaductisthcn'
lajor
part.the shapesofthe prepared pro-
iectbeingmandatorl'
.
''
'(llE)l 12.-' 'q -è Jk t7
' 13,' 17',4 t;l 1(. 1 '1 Two alternatives had been developed:
one in prestressed concrete.the other
7 -''F ll1j
- ';-5jp -.
:n() 7-i1J ')-!() 7-!4) ')'n() )qt
-) 2j.
';4.
$.'
( with an orthotropic box-girder deck
I (Fig.504.Both solutions have almost
I . thesameshapes.adapted to thespecif-
ic conditions of cable-stayed bridges
Nvith nnultiple spans and to the N' ery
strong winds 1()()or l5()n' laboN' e the
plateau.The shapes that have to be
respected were developed in close c0-
operation wit1 -
1 the architect to effi-
' cientlq'distribtlte rigiditq'buttu'een thtt
:kt)(1
dit-
ferentstructuraln' len-
lbers.butwith
' 7
I the greatest care for the bridge ele-
7
1
l
1!
1
,
1y()(j lI gance and for the aesthetic coherence
ij' ll,.
'
Ft)t) . ofthe differentelenRents.seeking ap-
!
j
i '
j
' parent simplicitl'and expressing the
.h j
A
. A
* ez
reallow offorces.
!
1 I
l I Tlne deck hasa trapezoidalshape.be-
1
j ing alluost triangular with a nal -l'
ow
!
; j lower slab.' The triangular shape that
i had been contenAplated was elim inat-
l 2
1 edbecatlseofitspooraerodq'nalnicbe-
$ l haN'iour.Conlpared ïvith the prelin- li-
l
l j e'
=I .
' nary design.the deck ism ore slender.
' l :.$
about4.50 n- 1conzpared vvith 5.5()nz.
1 j - ...' w
--
=-
-'!1
w- wt
wc x.I
1r.r
The pylons are 90 m high and art?in
'
j I ;
l
I u r'..r the bridge axisforelegance and struc-
i l .
'
,
p 1
I ttlralptlrity.-l-
heyhaN' e theshape ofan
l
i
; i
I 1 inverted N'longitudinalll'to produce
I I the necessarq'rigidity and to appear
p
!j
!
,9.
1
:),:j' lightand transparentatthesalne tim e.
14.531 Thedesign ofpiersbelon' 'had to adapt
to contradictoo' requirenaents: the
tallest ones haN' e to resist verl'high
Nvind forces-w hereasthe extren' le ones
have to be flexible wit1 -
1regard to lon-
gitudinal forces. For project h()mtl-
geneity the architectpreferred to give
a1l the piers the sam e shape-wit11 a
. :. ' . > B'ide box section in the lowe1'partof
+'
> >y, the high piers and a division into tu'
îl)
. ..
*N *>vM-..
.< XAW'A'
. .. 'd
'N' -
%,w.
X
Q'Nx
wx .
.
- . shaftsinthehighiarpart-about9f)m higll.
e.. ,t W <Y.Y- >
w *
x x
x . .
The candidates for the concessitln
u'ereselected in200()and arepresently
.'. .k.
..a.
. preparing theiroft-ers.lfthis bridge is
N NN ''. erected-asexpected.itu' illhavea nna-
. ..
. V '-. >
.
jorimpactontheproft
assion dueto its
W
.. . .' technicalinterestand itsarchitectural
z..#a/
.; ' # o' v. '
' I
. .
Z ' '
&.
,
.
$
perfection.
.

- u111
..
1
X
,
:
'
u
.1
!
:.'
.'''.
c
l
h
i
'
.
' .. ...
' .
.'
1 4
<l
W
.
.. .
..

.
.tr..r .. .. ' .. :;
. -.
sy'..
''

StructuralEngineeringInternatillnal 1-/2001
'

t) 1 O 3 7 6 ? (7
am ending the shape of the offshore
î !
a46 ).f)f) 1 l
( structures that constitute the founda-
u(
)-1.d)() 3-;*
--.()() 3-T7()() 3-:7()1
') 7-to4-8() 3:1.-D..(:J Jl-)n
-..(J() -.()-).()t)'
% , tion caissonsand the piersin orderto
f r
I f redtlce the m assofwater accompany-
I 1 k ë j
' 30254
?7, l I 1 1
' 1 l ing the caissons during earthquakes.
H e entrusted the authorwith an audit
1j
. .
$
r.
t !!
l;
:j!
l;
yt I
t! j of the proîect.asking for proposals
i
hl. i to improve it. Retkrring to Pierre
E
'
j ,
Richard-s project for the Ré lsland
Bridge. the author recom m ended a
continuous deck totally suspended
from thefourpylons.rigidlyconnected
tothepiersbelow'(Fig.51).'Fhissolu-
tion wasadopted im mediately.theside
spans being shortened slightly so that
the last cable stays almost reach the
end supports.
ln comparison with theinitialproiect.
thissolution hasm any advantages.Py-
lonsare connected directly and rigidly
to the piers below. and the relative
m ovem ents betw een piers and can-
tilevers.which appeared so question-
Rw
.
<w able.are thus eliminated.The deck is
è
;m
' xw>''
'
u
X
.'
777
.
y
'.x.
x
x'''
t.wx
''''' h
.h
continuous with no expansion joint.
,-.'',- qf.ï
: -.. '. '. ..
.
:t
:ty . x.
.
....
..
''' ,', .-:.9'.
-
-
.' 'l
'..
.. 's''
l' shh
. .'''
's
' xs.
ss q. producing greater comfort.Safety no
.,',,..,
,,. ..'....s ....
. ..
longer relies on dam pers.buton the
ducti'
-ity of the structural m em bers.

tX
'!
$.
'
1
IC
1
c
'
>
&
F
m
2
,
;
j
I
-
+
)
r
ë
i
m ainly the pylon legs.on which itis
i
.i easierto be confident.Some dampers
t( are used only to lim ittransverse seis-
m ic m ovem ents. Since the drop-in
spans have been elim inated. cable
stayscanbeuniform ly distributed.and
bending momentsdrastically reduced.
and no m ore '-pancake effect-'has to
be feared.
The finalprojecthasbeen developed
on these bases by GTNI and Ingerop.
The bridgehasfive spans.286.3 )' t560
and 286 m long. 7-he four supports
keep the aspectofthe offshore struc-
tures.with a directfoundation on the
sea bed extended by a pierhaving the
shape of a large circular caisson.The
piercrown widensthiscaisson to leave
passage for the deck and to allow the
Rion-xntirlon Bridge The idea thatcantileversw ould m ove installation ofthe fourlegsthatconsti-
by at least 1.00 m during extrem e tuteapylon.joiningatthepyl ontopto
A lastproiectdeservesattention.espe- earthquakes was not very com forting have (he shape ofan inverted V longi-
cially asitisthe singleone undercon- w ith the drop-in spansbetween them . tudinally and transNrersally.The deck is
struction today.nam ely the Rion-A n- The concentration of several cable a com posite structure with two steel1-
tirion Bridge thatwillcrossthe Patras stays at the cantilever ends.in order girdersasedgebeamsjoinedbymulti-
Bay.The initialdesign presented pre- to balance the weight of the drop-in ple floorbeam sand a reinforced con-
viously in this article raised m any spans.produced undesirable bending crete slab.This solution was selected
questions. m om ents.m uch higherthanin classical forits low ercost.despite the aerodp
A tfirst it was questionable.as struc- cable-stayed bridges.Further artificial nam icbehaviourofthistypeofprofile.
turalsafety relies com pletely on a se- problems are produced when cable which isnotthebest.
riesofhuge dam pers.w hich do notyet stays are anchored at close intervals.
exist: the possible tectonic displace- Finally. vertical effects of extrem e
mentsreach 2.00 m and seism ic forces earthquakes would shake the drop-in C onclusions
arevery high.Furtherspace necessary spanslikepancakesin theirpans.
to house al1this equipm entappeared This extensive review presents a new
lacking-despite the large dim ensions W Then Jacques Com baulttook charge
field of application of cable-stayed
given to the pierheads. oftheprojectsupervision.hebeganby bridges.The erection ofthe Rion-A'n-
StructuralEngineering International 1/2001 Reports 81
TN'IT-FFOHT.H .I''illlll;)llt/tpp'.b';)tl1lIîI$'t:jJ(t Be-
El?5c
. K -'t
hs
'('
7 --'
.'
:A() :1
-($(.1 3%
-'()(.4 -7i>''
- tt
3 -5-1..u84
-
tonN'e1'lag.DiisseldorI'.l97'''
. k .
' 1 .
. 1 I I
PO DO LN 5-.54' ..S(7.ALZI.J.('()/7. $'r77/t'r?'ttlltf?lt/

bt,7 s
x.4t.. ar'it
ass.'p,.t' .k.k
. .s/.
,...
.t , .'.
. p. ' . - .-- k j.
f t tlesizll t?/'t'
fk/i/g-xff?y'
t,t/ 1
)l'
i(lq:L'
.
b.Jt'h1 ïï' i1t
?1'ck
=- R i == L' ' ' ''
Son>-Neu 5-ork.1976.
(..'
iI5ISIN G N.J.Ct//l/f.?. sLq)))()t':ç'(I/?r,'J.vtts.'(-()tl-
t't.
gplfllltiflesigll.John '
iNliltl'qs'tt'Stlns.C-hicllestel'.
l983.
NN'ITTFO i'1T.H.Britlles Betonxkrlug.Dtissel-
dorf.l984.
. k;'
!#xz
;x
yi' ff..z .'4 u.. '- . .a. .p SI.'-..N.L7-H ER.R.etttI.P(t)1I.b/?tJl//7dI/?f7$.Pl'ttssch
z.. :672;:1; ,.'' .>' .
vqr .
wox.
wx
u-- -v
>'
'r 'r5r?'.?..'s..t L.t'.%..XM.'xNwuxx-n.x.hhxhx'wx>-.
y
)/z
i.
/jvs
y''
w-x
>-N.
.
x
-..
-- '
, .
,.'
,
m
''
K,
Az<).'z
Z r
z: ...'
.
j'
s
.t.q x
h
kM x't.'x-
x7xzwN'
h'w:: -
%
xt-
L
r
-
w
rx
w-?s'N'>N.-'''''NX'.X. .
-
xM
Jzz'
x'
-
w-'
. ..e'
z,j.='v
.
VA...''Zz
--
.-
,'z
'..<.z'''.,.sz,,.;
.
'n.
-----
'N<---x
/, .' !h, .k xNx .
ovow-
A..''A'
.' .
.-'
z.,.'
,,z'.
..'.
J,/;
/ :
; ht
N. L
'

.
.x
'xl
: xL
N
... .' !.,'.%'x'xy.xx lx
èx x xvr
-.
. X x
Pol.l!cchniquesRom êtlldu-s.l-ausannt).l9S6.
.

' -
.l -- - . - - - .- - ..-- .- . 1 . . LEO NH .ARDT.F.P()?In'.Pltellrcx.Prutssk!sPoly-
@ =
lcchniquesRonnandes.Ltttlsal)Ile.ltpst).
TRO ITSKY.51.S.C7?/?/t?-..'JtIï't't/I,l'i(lqk,.b.2ndedi-
tion.BSP Professio1' 1alBooks.()xford.1tl8:.
5IO ItAND I.R.blllLJb'Ll.zi(,
tI.
lL,.T@,t'/!/st?/fl,zitl./h'()g-
tte?J't?.G angenRi.Rol'
na.J9t)l.
FR ES'SS1NET.lhtblc-3r/dl.'t,tïbl.l'tific.%.X.'tkli)r).'-N'i1-
lacoubIaj',1994.
Tlle ??t?It' .b1a('(lll-'
/1//'/'tf B''i(lzçt.' J/2(, t''l'k.'llLi.%lll'il
.
/l/'/t-/tttl.Pol'talld BridgcOffice.s'lacau.l994.
'
HO LG .A.TE.xA. '.T'
/!(,6î'()tk t?/'./t'j/-t.
r.%L'/?/t!/'t'/?iltlfl
&xt
/1is rt/tlp?1.-.' ':tNlendes.S!-uttgLtrtttntlLtlndon.
1t.
)97.
h.-IRLOG Et.;X N'1.l%)llI.àtàFlt/pfAflll.b éè J,'t?!'(;(?.s'
'
tJltlliiple-
b.Progre-s dans ICt conception tlu's ou-
N'rages d'arten béton.Jtlurndes d'dtudqls SlA. '
G PC elll'lponl'leurdu 65'''ktllniycrsllil't.
tdu Pr()-
fesseurRtc11audFavre.S1. A D()1(''t).Zurich. 1999.
SCH 1-.-AICl-1.J.(-'(???('t,/?J//tl/Dtz.
sf/kz/?t?/'Sr/f/kztu.'
î./t)lrg tk'
lri(,l'j'.'Bridj'
tttEngineering C-tlntk-rklncu'.
Sha1-11-1ElSheik11.2()t)().
tirion Bridge.and hopefully'ofthe(
''
N1il- R eferences SCH LAlCH .J.î'tll'iLAlb'i1lJ-l/'jt/jxcDc5jk,/!.TrtlIluls
lau '
Viaduct.willshow the considerable 1I
A Bridge Design ((l't'nfel' encd.).Nlctd1' 1d.- n()()
t).
inlerestin this tlrpe ofsolutioll.espe- Tllç
, /p,.
J
'
(tve
- xpilllllitlg /-(?/v.c sltllzlctlillo /?3
cially 1n larger proiects like the Fern- 1.'
i
f
'lle-lleltl.Bauïerlclg-Beriin.l963. . Y'IRLOG E'L-X.N1.Bt'itlgtA.b
;IT'/J/?.h.lttltilll(..s (,
(1-
beltBridge between GernAany and the BoaxsA.c).:Boxl.G.z7?t,(.r???(.?.(,f(,t?r(./;j?(,(.- /?/(
a-5'
. fffb'
ctl5>t???y.Trk?1
)dsi:
1BridgeDesigl'
l(Ckln-
Copenhagen lsland. ?!/?.t
-()fRiL.L'a/xl().%Jt/ré?,7f//'.Alttc'Tiranti. lk ren ce ).N'
1C
ldr ik
.l.
-76)
(1
().
.

Repol'ts StructuralEngineering lnternational 1/2001

You might also like