Sei sulla pagina 1di 15

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235637604

Construction stages of cable-stayed bridges with composite deck

Article  in  Bridge Structures · February 2013


DOI: 10.3233/BRS-120044

CITATIONS READS

17 6,720

4 authors:

Michele Fabio Granata Piercarlo Margiotta


Università degli Studi di Palermo 19 PUBLICATIONS   96 CITATIONS   
62 PUBLICATIONS   243 CITATIONS   
SEE PROFILE
SEE PROFILE

Marcello Arici Antonino Recupero


Università degli Studi di Palermo Università degli Studi di Messina
56 PUBLICATIONS   213 CITATIONS    62 PUBLICATIONS   379 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

In.CAM.M.I.N.O. View project

All content following this page was uploaded by Michele Fabio Granata on 31 January 2015.

The user has requested enhancement of the downloaded file.


Bridge Structures 8 (2012) 93–106 93
DOI:10.3233/BRS-120044
IOS Press

Construction stages of cable-stayed bridges


with composite deck
M.F. Granataa,∗ , P. Margiottaa , M. Aricia and A. Recuperob
a Dipartimento di Ingegneria Civile, Ambientale, Aerospaziale e dei Materiali, Università di Palermo, Palermo, Italy
b Dipartimento di Ingegneria Civile, Università di Messina, Messina, Italy

Abstract. A cable-stayed bridge is built by following a construction sequence of different subsequent stages that imply changes
of the geometric configuration, restraints and consequently stress and strain patterns. A study is presented about stay stressing
sequences in cable-stayed bridges with composite deck for the case of construction made by the cantilever method. Initial cable
forces are found through partial elastic schemes of construction stages. Different stay stressing procedures are implemented
and a comparison between them is proposed in order to evaluate the differences in terms of deformed shape and internal force
distribution of the deck and the pylon. Results can be useful to give practical suggestions to designers, when they have to deal
with stay stressing sequences of composite cable-stayed bridges.

Keywords: Cable-stayed bridge, composite section, stays, stressing sequence, pretension

1. Introduction operation leads to a fundamental difference between


concrete and composite decks. In concrete deck dead
The deck of cable-stayed bridges can be of three dif- load is applied at once and the stay is stressed by refer-
ferent typologies: prestressed concrete girders or boxes, ring directly to the entire value of the self-weight, both
steel-concrete composite cross sections and fully steel for the cast in situ technique and for the prefabricated
sections. This variability is mainly due to the wide range segments. In composite structures instead dead load
of span lengths covered by these bridges, from 100 is applied in two subsequent times at each phase of
meters to 1000 meters, as well as to the costs of the construction. So the cross section, related to the new
different solutions adopted among these typologies. segment built, changes; in fact it is made only of steel
Prestressed concrete decks can be advantageous for in a first time and it becomes composite in a second time.
the smallest spans, while composite sections are com- As a consequence the evaluation of initial cable forces
mon for medium spans and orthotropic steel plates for and the following stressing sequence of stays are dif-
the longest spans [11, 19]. When the bridge is built ferent for concrete and composite decks. Moreover, in
by the cantilever method, the deck typology can have concrete cable-stayed bridges, the determination of ini-
different consequences in term of stress and strain pat- tial cable forces is mainly related to the aim of obtaining
terns and over all in the determination of initial cable the desired geometric profile of deck and pylon, by tak-
forces and stay stressing sequence. The steel-concrete ing into account the fundamental role of time-dependent
composite deck is built in two steps: first the steel ele- phenomena in concrete such as creep and shrinkage. In
ments are assembled by cantilevering to the previous the case of composite decks, the main aim is to obtain
segment already built and then the concrete slab is cast a convenient bending diagram for the dead load config-
over them and the stay attached and tensioned. This uration, associated to the required deck shape, in order
∗ Corresponding author. M.F. Granata, Dipartimento di Ingegne- to minimize stresses into the steel elements and the
ria Civile, Ambientale, Aerospaziale e dei Materiali, Università di concrete slab, avoiding over dimensioning of structural
Palermo, Palermo, Italy. E-mail: michelefabio.granata@unipa.it. elements and concrete cracking.

1573-2487/12/$27.50 © 2012 – IOS Press and the authors. All rights reserved
94 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 1. Non-satisfactory geometric profile of a cable-stayed bridge.

Nevertheless the correct and desired geometric pro- authors to concrete cable-stayed bridges [3] and to arch
file of the deck is a hard target to be achieved; some bridges built by suspended cantilevers [15] or by lattice
cable-stayed bridges present a non-satisfactory shape cantilevers [16]; it is developed here for the specific case
of the deck (Fig. 1) just in the early phases of service of composite decks, for which the construction stages
life, due to the stressing sequence adopted during con- imply a different stressing sequence from those of con-
struction; after many years the problem can grow due to crete bridges. This study reports a comparison between
the effects of creep and shrinkage. A study of different different stay stressing sequences in order to evaluate
stressing sequences, as the one presented here, can help the consequences of each one to the state of stress and
the designer to avoid the bad consequences of a wrong to the final geometry of deck and pylon. The main target
methodology in terms of deck geometry and internal is to detect the weakness and strong points of each con-
forces distribution. struction methodology and to give useful indications
Another important aspect to be considered in to designers in choosing the most convenient method
steel-concrete cable-stayed bridges is the increased of initial cable force evaluation and stressing sequence.
importance of cable sag nonlinear effect with respect to The analysis aims to obtain a predetermined geometric
concrete bridges. It depends on the span range covered profile of deck and pylon at the end of construction with
by composite bridges and the related length of stays, for acceptable values of bending moments. The chosen ref-
which the geometric non linearity has to be considered erence configuration for the deck is that of an equivalent
in many cases as well as the second order effects on the continuous beam on rigid supports, which is approxi-
deck and the pylon [2]. A steel bridge emphasizes these mated not only in the final configuration but also in the
two aspects: the optimization of bending moment dia- intermediate ones during construction, by performing a
gram in construction stages and in the final dead load convenient stay stressing sequence. Different hypothe-
configuration and the effects of geometric non linearity ses were considered for assembling steel elements and
[13, 24], often influenced by cables lateral motion [26]. concrete slab in the cantilever configuration, in order to
Some related topics were underlined by Schlaich [25], compare the results obtained by the analyses.
giving useful indications to designers about the con- About time-dependent phenomena in concrete, creep
struction of cable-stayed bridges with composite decks and shrinkage play an important role in cable-stayed
and prefabricated slabs. bridge construction by modifying the state of stress
In this study a methodology is proposed for the evalu- and strain in time. For concrete bridges a minimiza-
ation of initial cable forces in composite bridges, based tion of creep effects can be achieved by performing a
on simple partial elastic schemes of construction stages. particular stay stressing sequence and by choosing an
This methodology has been already applied by the opportune value of initial cable forces approaching the
M.F. Granata et al. / Construction stages of cable-stayed bridges 95

behaviour of a continuous beam on rigid supports [3, result. It is due to the high grade of redundancy of
7]. In composite bridges time-dependent phenomena cable-stayed bridges (mutual effect of stays), to the real
have two main effects: the first one is the axial short- sequence of load application and stay pre-stressing and
ening of concrete pylons and deck slabs; consequently finally to the influence of time-dependent phenomena.
an increasing downward displacement of the deck with Moreover when the backward methodology is applied,
time and a danger of slab cracking occur. The sec- during the operation of bridge dismantling the stress
ond one is the variation of the composite cross section evaluation is made on a configuration that is not in a
centroid, as expected by the creep-transformed section zero-stress state because it has been already loaded,
theory [10], now inserted in final documents of ACI while in the actual sequence every segment and every
209 Committee [1]. It implies a change of the state of stay to be added is in a zero-stress state before they are
stress in time, particularly of the bending moment dia- assembled. As a consequence it is difficult, especially
gram, with a redistribution of stresses between steel and for cable-stayed bridges that are not completely self-
concrete elements in the section [8]. It is a fundamen- anchored or unsymmetrical [18], to obtain the required
tal aspect that cannot be denied because it leads to a geometric configuration with a single stay stressing but
significant modification of internal forces. it is necessary to adjust the value of axial forces for all
In the following, after a brief analysis of stay or several stays. It implies a technological problem and
stressing sequences and a literature survey, a com- a related cost, because for each adjustment of the stay
parison between three methodologies of stay stressing force, the stressing equipment has to be shifted in the
is explained and discussed, with the determination of position of the right stay. Moreover the result of many
initial cable forces based on elastic schemes of con- adjustments can be a reduction of the safety degree
struction stages. Two numerical examples of composite against fatigue because for every stressing operation
cable-stayed bridges are presented in order to make the cone of anchorages engraves the stay in different
clear the followed methodology and to underline the points, damaging steel wires and reducing the fatigue
consequences of each procedure in terms of internal strength [21]. For this reason the aim is to reduce adjust-
forces and deformations. ment operations by making only the essential ones and
by engraving the steel wires in new areas (i.e. limiting
the operations of stay release), in order to maintain the
2. Stay stressing sequences in cable-stayed required fatigue strength of stays.
bridges In order to take into account all these features, the for-
ward analysis is a useful and wide-used methodology.
Literature about stay stressing and initial cable forces It allows designers to consider a number of important
determination in cable-stayed bridges is very complex aspects, as the following ones, now implemented in
and wide, because this topic has been considered by specialized finite element software packages:
many authors and by different point of views. About
cantilever construction the most frequent methodology – when a new segment is built by the cantilever
to find the initial cable forces is the so-called back- method, by assembling it to the previous one
ward analysis, which is a deconstruction analysis of the already completed and by stressing the related
bridge. It starts from the desired final stage and then stay, it is necessary to choose the right posi-
the bridge is ideally dismantled stage by stage [6, 14]. tion of the new segment by following the tangent
In this methodology initial cable forces are found as direction with respect to the tip of the previous
the forces in the stay at each deconstruction stage [4]. segment, already deformed (Fig. 2). It avoids the
Unfortunately it can give to designers only the initial birth of imposed deformations between segments,
value of the stay force and for this reason it always has minimizing the geometric differences with the the-
to be followed by a forward analysis which considers oretical deck profile.
the actual construction sequence, giving the actual state – When a stay is attached to a new segment,
of stress and strain of the bridge structural elements. correctly positioned and tensioned, the real
Even though values of pre-stress to give to all stays are length of the stay differs from the theoretical
found by this methodology, the final stress state is gen- one, due to the displacement of the anchorage
erally different from the expected one because a unique point in the deformed configuration (Fig. 3). As a
stay stressing operation, based on the value found by consequence the stay force at the beginning of its
this methodology, doesn’t permit to obtain the desired life is different from the expected one, foreseen
96 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 2. Tangent position of the new segment assembled.

Fig. 4. Lock-in curvature. a) Cantilever construction of the whole


composite section. b) Steel cantilevering and stay pre-tensioning
before slab casting.

assemble the steel girders. In this phase the stay can


be tensioned or not (Fig. 4). The casting of concrete
Fig. 3. Different stay length with respect to the theoretical one due is then made on the steel beams already deformed
to the segment deformation.
and the stay is tensioned at the design value. Dur-
ing this last phase an internal curvature remains
in the undeformed analysis model. So a force locked into the deformed steel element (lock-in
variation needs to bring the anchorage point in the curvature): it implies a variation in the distribution
correct position (lack-of-fit force). and value of bending moments due to the change
– The stay is not a rectilinear cable between the of the deck cross section, which is made only of
anchorage points of deck and pylon, so a sag effect steel in a first phase and then becomes composite.
has to be considered which implies a geometric
non linear behaviour [14]. This effect can be The most known and used procedures to obtain the
neglected for short stays; generally under 300 initial cable forces, both in the cases of backward and
meters, the differences between linear and non forward analyses are the following ones:
linear theories are not significant but it can be
important for longer stays [28, 29]. 1) the deflections of stay anchorage points in the
– Forward analysis is the only one that can take into deck and the horizontal displacements of the
account the time-dependent phenomena of con- pylon points are nullified by the so-called zero-
crete (shrinkage and creep) on the base of the pre- displacement method [27];
diction models given by codes and specialized lit- 2) A convenient and pre-determined distribution of
erature (fib Model Code 2010, ACI209-92, Bažant bending moments in the deck and the pylon
B3 model, GL2000 model). It follows the actual is obtained by applying the force-equilibrium
sequence and time of concrete casting related to method or the unit load method [5, 17];
the construction stages by considering the stays as 3) An opportune energetic function is minimized
elastic restraints in a non-homogeneous structure [30].
[20] with a variable structural system [9].
– If the cross section is a steel-concrete one, then it is Recupero [23] proposed a mixed method for cable-
composed of different elements assembled in dif- stayed bridges with steel deck which considers a
ferent times. When the new cantilever segment is pre-determined geometric configuration and a contem-
attached to the previous one, the first operation is to porary optimization of the bending moment diagram.
M.F. Granata et al. / Construction stages of cable-stayed bridges 97

In the following paragraphs different stay stressing scheme, by separately evaluating the cable force
procedures are applied to composite bridges and com- needed to recover displacement δI of the steel
pared in order to establish which is the most convenient elements and that needed to recover displace-
sequence. Numeric examples are also presented and ment δ I due to the concrete cast. The superscript
discussed. B of forces points out that initial cable forces
and adjustments are different in every approach
3. Partial elastic schemes of construction stages followed.
in the evaluation of initial cable forces C) Steel elements of the cantilever segment are
assembled and the new stay is attached without
In this study only the forward analysis of cable- modifying its length, as in the first approach. The
stayed bridges built by the cantilever method is taken slab is cast and then the stay is stressed to the
into account. Initial cable forces determination is done desired value by recovering both displacements
by partial elastic schemes of the structure, one for each δI and δ I . At the same time the previous stay,
construction phase, i.e. for each segment assembled by already stressed in the previous stage, is adjusted
cantilevering and for each stay attached and tensioned. in order to compensate the displacements δI-1 and
When steel elements and slab casting are built in two δ I-1 of the related anchorage point, due to the
following times, different partial elastic schemes of the actual construction stage (Fig. 7).
two subsequent assembling phases have to be consid-
ered for each stage. The latter two-phase stressing operation is proposed
The three approaches here investigated are: because it permits a more precise compensation of
displacements due to construction loads, achieving a
A) steel girders are assembled and attached as a can- configuration near to that of a continuous beam on
tilever segment to the previous one completed rigid restraints in every phase of construction. It was
and suspended by the related stay; the new stay already proposed for concrete cable-stayed bridges to
is attached to the girder of the new segment and compensate creep effects in service life [3]. The disad-
its length is not modified. Then concrete slab is vantage is that each stay is stressed two times because
cast in situ and/or built by prefabricated elements; it implies the stressing equipment is shifted from one
at the same time the new stay is tensioned in a stay to another at each adjustment or that two equip-
single operation at the design value, which corre- ments have to be employed. The evaluation of initial
sponds to the zero displacement of the anchorage cable forces is done on the elastic scheme of the related
point. The pretension value to be given to the stage, by considering contemporarily the displacements
new stay attached is found through the partial of the last segment and those of the previous one
elastic scheme, by making null the vertical dis- (Fig. 8).
placements due to the initial deformation of steel In all the three approaches investigated a final reg-
elements added to those due to the slab casting ulation of all stays is convenient in order to take into
(Fig. 5). The problem of cable force determina- account the displacements due to construction errors,
tion in every partial elastic scheme is solved by superimposed dead loads (pavement, guardrails, foot-
the commonly used procedure of the influence steps, etc . . . ) and creep effects in construction stages
matrix of displacements [15, 23]. [22]. Moreover the effects of shrinkage and creep till
B) Steel elements of the cantilever segment are construction end can be partially compensated by a final
assembled and the new stay is tensioned a first adjustment of all stays. In fact the axial shortening due
time to a value that recovers the vertical displace- to concrete shrinkage, especially in the tower, implies
ment of the steel cantilever. The concrete slab is an increased vertical displacement of the deck with
then cast and the stay is newly stressed to the respect to the result of theoretical analyses. In order
definitive value which recovers all displacements to compensate the pylon axial shortening, it is possible
(Fig. 6). In this case a double stressing operation to build concrete elements of an increased length [25].
is done for the last stay attached; this method- About the flexural effects of creep and shrinkage, it
ology can be helpful to minimize the lock-in has to be considered the behaviour of composite beams
curvature of steel elements. The value of ini- that are non-homogeneous elements with respect to
tial cable force Fi and that of adjustment Fi creep. In fact these elements show an internal redistri-
after slab casting, are found on the partial elastic bution of stresses between concrete and steel members
98 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 5. Approach A. Single stay stressing.

Fig. 6. Approach B. Two-times stressing of the last stay attached.

Fig. 7. Approach C. Two-phase stressing of the last and the previous stay attached.

and the change of the centroid position of the so-called moments appear in the deck [8]. As a consequence of
creep-transformed section, which is the homogenized this variation of girder characteristics, internal forces
section modified by the creep coefficient (Fig. 9). and deck deformations vary with the time, even though
In a first phase, during the cantilever construction, the the behaviour of a continuous beam on rigid supports
cross section centroid coincides with that of the steel was achieved at every construction stage.
element, while after the slab casting, the homogenized Another consideration has to be done about the
section has the centroid in a new position. It implies anchorage point of the stay to the deck; when it is set
an upward shifting of cross section centroid and as a at the upper fibre of the deck section or in an eccen-
consequence, positive bending moments appear in the tric position with respect to the cross section centroid,
deck due to the eccentricity between the two centroid a concentrated value of bending moment is generated at
positions (before and after concrete casting). In ser- each anchor point, which can be significant (Fig. 10). In
vice life instead creep implies that the centroid shifts this case a different distribution of moments has to be
down (toward the steel member); so negative bending expected with respect to the continuous beam on rigid
M.F. Granata et al. / Construction stages of cable-stayed bridges 99

Fig. 8. Elastic schemes for approach C. a) self-anchored stage; b) partially earth-anchored stage.

4.1. Cable-stayed bridge with two pylons

A cable-stayed bridge with two pylons is presented


with a central span of 210 m and two lateral spans of
60 m. Stays are arranged on two plans with a mixed
harp-fan configuration; it is a partially anchored bridge:
for half bridge there are 21 stays and the last five
backstays are anchored to rigid piers. The bridge is
longitudinally and transversely symmetric (Fig. 11).
Distance between stay anchorages is of 10 m on the
deck and 2 m on the pylon. Self-weight of the deck
is qd,s = 100 kN/m for steel elements and qd,c = 150
kN/m for concrete slab. Superimposed dead load is
Fig. 9. Change of the centroid position due to construction stages qs = 60 kN/m. Cross section of the deck is a steel box
and creep. a) Steel element. b) Homogeneized section. c) Creep- with an upper concrete slab rigidly connected to the
transformed section. box. Transverse stiffen diaphragms are located every
five meters. Wideness of the deck is 20 m while its
supports; this fact is generally amplified by the effects depth is 1.20 + 0.30 m. Geometric data are: A = 1.27
of time-dependent phenomena in concrete. m2 , J = 0.397 m4 , yG = 0.465 m (centroid position of
homogenized section). Stays have an equivalent steel
diameter of φs = 0.12 m, while earth-anchored back-
4. Numerical applications stays have diameter φb s = 0.18 m. The mean section
of the pylon has At = 10.5 m2 and Jt = 10.8 m4 . The
Two examples of cable-stayed bridges are presented requested geometric profile of the deck is a 2nd order
here, with the aim to show the numerical results of the parabola with a maximum height at the midspan of
three procedures applied to the analyses of actual cases, 1.60 m with respect to deck ends. Concrete elastic mod-
and to compare these results. Two different configura- ulus is Ec = 30 GPa, while the deck steel modulus is
tions of the bridges have been chosen: the first one is a Es = 200 GPa and the stay high-strength steel modulus
symmetrical cable-stayed bridge with two pylons and is Esa = 190 GPa.
two plans of stays attached at the borders of the compos- The three approaches considered in this study, have
ite deck. The second one is an asymmetric cable-stayed been applied to a finite element model of the bridge with
bridge with a unique diamond pylon and one plan of 221 joints, 136 beam elements and 84 truss elements.
stay attached at the centre of the deck. Initial cable forces were found by the partial elastic
100 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 10. Eccentricity between the cross section centroid and the stay anchorage.

Fig. 11. Geometry of the bridge.

Table 1
Results at the construction end for the three approaches
Approach A Approach B Approach C
max M min M δvert δhoriz max M min M δvert δhoriz max M min M δvert
[kNm] [kNm] [cm] [cm] [kNm] [kNm] [cm] [cm] [kNm] [kNm] [cm] [cm]
16050 −9270 −11.94 −0.37 18200 −14900 −8.36 6.01 4350 −5420 3.54 −0.50

scheme method. The related stay stressing sequences packages allows designer to take into account shrinkage
were implemented. A staged construction analysis is and creep effects.
performed by following the forward procedure. The In the approaches A and B only the last stay is
construction sequence is divided into 11 stages, each stressed at each construction stage: they result less
one composed of two phases (one for the steel ele- advantageous with respect to the third approach C, in
ments of the cross section and the other one for the which a two-phase stressing sequence is performed.
slab casting). Geometric nonlinearity due to the cable With this methodology the final bending moment dia-
sag is neglected, for the limited length of stays. gram approximates that of a continuous beam on rigid
Table 1 gives the results in terms of maximum restraints and this behaviour is achieved not only in the
and minimum values of bending moments at the final final stage but also in every intermediate stage during
construction stage, as well as deck deflections and hor- construction. It is a great advantage because bending
izontal displacements of the pylon. moments have similar values in all the different steps
Figure 12 shows the comparison between bending of the sequence. However it has to be considered that in
moment diagrams in the final dead load configuration, composite bridges the bending diagram of a cantilever
for the three methodologies considered. Effects of time- phase (in which only the steel members are assem-
dependent phenomena in concrete are not included here bled) has often the same significance of that related
because the aim is to underline the differences between to the phase in which the stay is tensioned, but with
the three stressing sequences analysed, even though a different sign: negative moments in the cantilever
a forward analysis performed by specialized software phase and positive ones in tensioning phases. The sig-
M.F. Granata et al. / Construction stages of cable-stayed bridges 101

Fig. 12. Bending moment diagram at the construction end for the three approaches [kNm].

nificance of this aspect will be underlined in the next tensioned than backstays when the double stressing
example. operation is implemented at each stage, without the
Figure 13 shows instead a comparison between adjustment of the previous stay. So the final geomet-
deformed configurations for the three approaches fol- ric profile is not correctly achieved in the deck nor in
lowed. It is very difficult to obtain the requested profile the tower.
with a small tolerance; approach C gives always the best An opposite behaviour can be observed for the
result. deformed shape obtained by approach C, with the tower
A significant difference can be seen between bent out away from the midspan. This behaviour is
approaches A (stay stressing at once) and B (double due to the adjustments of backstays (that are anchored
stressing of the last stay attached). The second approach to ground), because in approach C, at each stage,
permits to limit the lock-in curvature in the segment these cables are re-stressed, leading to a higher value
but it is effective for self-anchored bridges and much of backstay forces than those of midspan cables. In
less for earth-anchored ones. In fact, if backstays are fact, in approaches A and B only one backstay is
rigidly anchored on one side, the deformation of the re-stressed at each stage, while in approach C two back-
segments on the other side (in the central span) induces stays are adjusted at each construction phase. In this
a greater value of curvature when the stay pulls up the way, the direction of deformations may be understood.
last segment toward the tower. At the end of construc- The midspan deflection is due to the accumulation of
tion a downward displacement appears in the deck and errors in the staged sequence, because of the effect
the related deformed shape shows a maximum value of prestressing forces given in the theoretical elas-
of displacement in an intermediate position between tic scheme is different from those given in the actual
the pylon and the midspan. Approach A emphasizes stage of construction. In the actual sequence in fact,
this aspect, in some cases with the consequence of an the reference geometric configuration before each stay
upward displacement in the midspan section (Fig. 13). stressing operation is not perfectly adherent to the the-
Moreover approach B shows a high value of the tower oretical deck profile. Only when all stays are stressed
horizontal displacement at the end of construction, in every stages, a complete compensation of deflec-
due to the fact that stays in the main span are more tions could be achieved. Nevertheless, it can be noted
102 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 13. Deformed shape at the construction end for the three approaches [cm].

Fig. 14. Stay forces evolution in the staged construction analysis. Approach C.

that the final horizontal displacement of the tower cable forces almost constant after the adjustment. It is
found by approach C is limited, even though bending a good consequence of initial cable forces found by the
moments appear in the tower due to the adjustments of partial elastic scheme method, experienced also in other
pretensions. applications [15]. Moreover in the early phases, when
Figure 14 shows the stay forces sequence in form of the stay is stressed and adjusted, the value of forces
a graph which gives the values of stay forces along the increases sweetly without sudden variations, contrarily
entire process of bridge construction, stage by stage. to what occurs with a one-step stressing sequence.
It can be seen that the third approach (two-step stress- The final adjustment given to all stays permits to
ing) gives also good results in terms of stress variation compensate not only the additional loads but also the
into stays during construction, maintaining the value of errors accumulated in the actual erection sequence.
M.F. Granata et al. / Construction stages of cable-stayed bridges 103

Fig. 15. FE model of the bridge. a) Symmetrical stage of construction. b) The completed bridge.

Fig. 16. Bending moment at a generic stage. a) Steel elements cantilever phase. b) Tensioning phase.

After this final adjustment, cables can be grouted. While values of stay forces. Further developments about this
the amount of the final prestressing forces is computed topic will be carried on.
in the design phase regarding the results obtained by the
theoretical staged sequence, in the actual construction 4.2. Cable-stayed bridge with a unique pylon
stage a measurement of cable force and displacement
variations has to be done by monitoring the sequence. A second application on an unsymmetric cable-
So a live model in the working phases has to be imple- stayed bridge with one pylon is presented. The bridge
mented on the FE software to account for modifications has two spans, the main one of 200 m while the side
with respect to the prestressing values obtained by the one of 170 m with three earth-anchored backstays. The
theoretical design. These variations are due to the actual configuration is with a unique mixed fan-harp plan of
values of temperature on working site, additional or stays, anchored at the centre of the deck cross section
not expected loads and equipment tolerances in jacking and attached to a diamond pylon. The deck is 22 m wide
phases and to the accumulation of deflections for the and it is composed of a box steel section with depth
not perfect compensation of anchorage point displace- 1.50 m and an upper slab 30 cm thick. Self-weight
ments. The magnitude of final adjustments given in this of the deck is qd,s = 85 kN/m for steel elements and
numerical example varies from about 5 to 20% of the qd,c = 157 kN/m for concrete slab. Superimposed dead
initial cable forces. A statistical assessment of the pos- load is qs = 70 kN/m. Geometric data are: A = 1.71
sible errors on working site can be performed in order m2 , J = 0.976 m4 , yG = 0.417 m (centroid position of
to establish how much different the actual result will the homogenized section). Stays have an equivalent
be from the expected theoretical one, especially for the steel diameter of φs = 0.13 m, while earth-anchored
104 M.F. Granata et al. / Construction stages of cable-stayed bridges

Fig. 17. Bending moment diagram at the construction end.


Fig. 19. Deformed shape of the bridge at the construction end.

The behaviour of a continuous beam on rigid


supports is not completely achieved but only approx-
imated because the two-phase construction (cantilever
steel elements and successive slab casting) implies an
upward shifting of cross section centroid as shown in
Fig. 9a–b. The consequence is that positive moments
Fig. 18. Bending moment diagram at t = 10000 days.
grow with respect to the continuous beam, by using the
zero-displacement method. But this phase can be con-
backstays have diameter φbs = 0.18 m. The section of sidered a transient one, because the effect of creep in the
the highest part of the pylon has At = 15.7 m2 and non-homogeneous composite section is in the opposite
Jt = 16.1 m4 . Figure 15 shows a view of the finite verse and negative moments grow, after the cross sec-
element model of an intermediate stage of construction tion centroid shifts toward the steel member as shown
and the one of the completed bridge. Bridge restraints in Fig. 9c.
in the main span, at its end, are simple supports made By considering the fib Model Code 2010 [12],
by conventional longitudinally sliding bearings. with a value of concrete strength fck = 30 MPa,
In this case only the third approach has been per- the relative humidity RH = 70% and the parameter
formed. Figure 16 shows the bending moment diagram 2Ac /u = 300 mm (fictitious depth), the creep coefficient
of an intermediate construction stage. When the steel is ϕ(10000,14) = 2.018; results of calculation in terms
elements are assembled by cantilevering, negative of internal forces redistribution are shown in Fig. 18.
moments appear with upper fibres in tension (Fig. 16a) The bending moment diagram at t = 10000 days shows
while, after concrete slab is cast and the related stay a very significant growth of negative moments. These
stressed, positive moments are prevalent with the lower values of dead load bending moment have to be added
fibres in tension (Fig. 16b). to those due to service life loads, with danger of slab
This change of the bending moment sign has to be cracking in the elements built in the first construction
considered with particular attention because it happens stages (near the pylon).
at every stage; the proposed stay stressing sequence Figure 19 shows the deformed shape of the bridge
shows a balancing of bending moments, giving similar after completion. The desired profile has been reached.
values for positive and negative diagrams. Through the Nevertheless, the effects of time-dependent phenomena
staged construction analysis it can be seen that values can be significant also in modifying the deformations,
of maximum negative and positive moments result of normally by increasing the values of deck deflection.
the same order of magnitude, stage by stage. Moreover They can be recovered, at least for what concerns the
maximum tensile stress in the concrete slab is about 1.0 effects of creep on construction displacements, through
MPa while maximum tensile stress in the lower steel the final adjustment of all the stays after the superim-
fibre is about 70 MPa. posed dead load is applied.
At the final construction stage the desired shape The maximum cable stress achieved in the con-
was achieved with a maximum vertical displacement struction stage is 500 MPa, for which an increment
in the deck of 3.27 cm and a maximum horizontal dis- has to be considered due to service loads; the total
placement in the pylon of 0.9 cm. Figure 17 shows the value is always lower then the maximum suggested
final bending moment diagram at the construction end, tensile stress (about 900–1000 MPa). Fatigue strength
before applying superimposed dead loads. of stays has been preserved by operating, in all stages,
M.F. Granata et al. / Construction stages of cable-stayed bridges 105

by stressing operations in which axial force increases, References


without relaxing any cable.
[1] ACI (2011). Analysis of Creep and Shrinkage Effects in
Concrete Structures. ACI 209.3R-XX. Final draft, March,
2011.
5. Conclusions [2] H. Adeli and J. Zhang, Fully non linear analysis of composite
girder cable-stayed bridges, Computers and Structures 54(2)
(1995), 267–277.
A study about the cantilever construction of [3] M. Arici, M.F. Granata and A. Recupero, The influence of
cable-stayed bridges with composite deck has been pre- time-dependent phenomena in segmental construction of con-
sented. The performed analyses follow three different crete cable-stayed bridges, Bridge Structures 7(4) (2011),
approaches to the problem of stay tensioning sequence. 125–137.
[4] P. Cassity and E.Y. He, Sequential stressing of cable-stayed
The determination of initial cable forces is based on bridges erected on falsework, IABSE Conference on cable-
the application of the partial elastic scheme method, by supported bridges, Seoul, 2001.
considering the cantilever phase of steel elements and [5] D.W. Chen, F.T.K. Au, L.G. Tham and P.K.K. Lee, Determina-
tion of initial cable forces in prestressed concrete cable-stayed
the successive slab concrete casting.
bridges for given design deck profiles using the force
The conclusions of this study are: equilibrium method, Computers and Structures 74 (2000),
1–9.
1) a forward analysis based on partial elastic schemes, [6] C. Wai-Fah and L. Duan, Bridge Engineering Handbook, CRC
in order to find the initial cable forces, is capable Press, Washington, 1999.
to achieve the two principal targets of reaching the [7] M.A. Chiorino, C. Casalegno and M. Sassone, Time dependent
desired geometric profile of deck and pylon and a effects in cable-stayed bridges built by segmental construction,
3rd fib International Conference, Washington, 2010.
convenient bending moment diagram. [8] L. Dezi, G. Leoni and A.M. Tarantino, Creep and shrinkage
2) Particularly a two-phase stressing, in which the analysis of composite beams, Progress in Structural Engineer-
last stay is stressed and the previous one adjusted, ing and Materials 1(2) (1998), 170–177.
[9] L. Dezi, G. Menditto and A.M. Tarantino, Viscoelastic hetero-
gives the best results, approaching the convenient
geneous structures with variable structural system, Journal of
behaviour of the continuous beam on rigid supports. Engineering Mechanics ASCE 119 (1993), 238–250.
3) The variations of the internal force diagrams and [10] W.H. Dilger, Creep analysis of prestressed concrete structures
the deformation state, due to the double phase of using creep-transformed section properties, PCI Journal 27(1)
(1982), 98–117.
construction (steel elements by cantilevering and [11] L. Fernandez Troyano, Bridge engineering: A global perspec-
concrete slab casting) and to the non-homogeneous tive, Thomas Telford, 2003.
steel-concrete section, can be very significant both [12] Fib (2012) Bulletin d’Information n. 65 – Model Code 2010 –
during construction and in service life, when time- Final draft, Volume 1, Lausanne, pp. 350.
[13] A.M.S. Freire, J.H.O. Negrão, A.V. Lopes, Geometrical non-
dependent phenomena of concrete are taken into linearities on the static analysis of highly flexible steel
account. cable-stayed bridges, Computers and Structures 84(31-32)
4) Although a two-phase stressing methodology can (2006), 2128–2140.
imply technological complications due to the equip- [14] N.J. Gimsing, Cable supported bridges, Concepts and design,
John Wiley, 1997.
ment shifting to the stays to be tensioned, a reduced [15] M.F. Granata, P. Margiotta, A. Recupero and M. Arici, Partial
number of stress adjustments gives always the best elastic scheme method in cantilever construction of con-
results in terms of internal forces and deformations, crete arch bridges, Journal of Bridge Engineering ASCE,
both for symmetrical and unsymmetrical geometric 10.1061/(ASCE)BE.1943-5592.0000396, 2012.
[16] M.F. Granata, P. Margiotta, A. Recupero and M. Arici,
schemes of the bridge. This is true for self-anchored Concrete Arch Bridges built by Lattice Cantilevers, Struc-
bridges as well as for partially earth-anchored ones, tural Engineering and Mechanics, Techno Press, in Press,
even though the anchorages of backstays play a 2013.
[17] D. Janjic, M. Pircher and H. Pircher, Optimization of Cable
decisive role in establishing the bridge structural
Tensioning in Cable-Stayed Bridges, Journal of Bridge Engi-
behaviour in construction stages and service life. neering ASCE 8(3) (2003), 131–137.
[18] T.Y. Lee, Y.H. Kim and S.W. Kang, Optimization of tension-
Numerical results presented on two different kind ing strategy for asymmetric cable-stayed bridge and its effect
of cable-stayed bridges can be useful for designers on construction process, Struct Multidic Optim 35 (2008),
approaching the problem of choosing a methodology 623–629.
of the cable initial force determination and the stay [19] J. Manterola Armisén, Puentes: Apuntes para su diseño,
cálculo y contrucción, Colegio de Ingenieros de Caminos,
stressing sequence in composite bridges. Canales y Puertos, Madrid, 2006.
106 M.F. Granata et al. / Construction stages of cable-stayed bridges

[20] F. Mola and F. Giussani, Long-term behaviour of cable stayed [26] P.H. Wang, M.Y. Liu, Y.T. Huang and L.C. Lin, Influence of lat-
bridges, Studies and Researches, Fratelli Pesenti, Politecnico eral motion of cable stays on cable-stayed bridges, Structural
di Milano, 24 (2003), 153–187. Engineering and Mechanics 34(6) (2010), 719–738.
[21] M.P. Petrangeli, The bridge over the Wadi Dib in Alge- [27] P.H. Wang, T.C. Tseng and C.G. Yang, Initial shape of
ria, L’Industria Italiana del Cemento, LXIX(743) (1999), cable-stayed bridges, Computers and Structures 46(6) (1993),
362–379. 1095–1106.
[22] M.P. Petrangeli, G. Traini, L. Evangelista and M. Della Vedova, [28] P.H. Wang, T.Y. Tang and H.N. Zheng, Analysis of cable-
The Cable-stayed Bridge over Po River: Design and Construc- stayed bridges during construction by cantilever methods,
tion, 2nd fib International Conference, Napoli, 2006. Computers and Structures 82 (2004), 329–346.
[23] A. Recupero, Initial Cable-Forces Adjustments in [29] R. Walther, Ponts Haubanés, Presses Polytechniques
Cable–Stayed Bridges: Effects of Parameters Uncer- Romandes, 1985.
tainties, Advances in Structural Engineering and Mechanics [30] R. Xiao, L. Jia, X. Song and H. Xiang, Influence matrix
ASEM’04, Seoul, 2004. method of cable tension optimization for long span cable-
[24] S.Y. Ren and M. Gu, Static configurations of cables in cable stayed bridges, IABSE Conference on cable-supported bridges,
stayed bridges, Structural Engineering and Mechanics 34(4) Seoul, 2001.
(2010), 545–548.
[25] M.B. Schlaich, Erection of cable-stayed bridges having com-
posite decks with precast concrete slabs, Journal of Bridge
Engineering ASCE 6(5) (2001), 333–339.

View publication stats

Potrebbero piacerti anche