This action might not be possible to undo. Are you sure you want to continue?
th
World Conference on Earthquake Engineering
Vancouver, B.C., Canada
August 16, 2004
Paper No. 3296
SEISMIC ANALYSIS OF THE INDIANO CABLESTAYED BRIDGE
Paolo CLEMENTE
1
, Mehmet ÇELEBI
2
, Giovanni BONGIOVANNI
1
, Dario RINALDIS
1
SUMMARY
This paper presents results of observed and analytical analyses of the dynamic response of the Indiano
CableStayed Bridge in Florence, Italy. The observed part is based on ambient and trafficinduced
vibration tests, which allowed extracting the dynamic characteristics of the structure in terms of resonance
frequencies, modal shapes and damping. These are compared with those obtained from a finite element
model for which the geometrical characteristics and the mechanical properties of the materials used in the
structural design of the bridge have been adopted. To supplement these analyses, equivalent static
analyses were performed to yield amplitudes of response and stresses in structural elements. This version
of analysis was performed by response spectrum technique using uniform response spectra expressed in
terms of pseudovelocities as input. The response spectra were obtained from contemporary, seismic
hazard studies performed in Florence site.
INTRODUCTION
Identification of dynamic characteristics of structural systems is an important step that is necessary for two
basic reasons:
− to assess possible dynamic behaviour of structures during strong events and
− to gain experience on the general dynamic characteristics of structural systems such that the
experience and data base can be used in future design and analyses.
Past efforts in determination of dynamic characteristics of structures has facilitated development of data
bases, which in turn has been used in establishing empirical formulas for estimation of fundamental
periods of structures in design codes. Also, critical damping percentages used during dynamic analyses are
obtained from past data bases. The data bases for dynamic characteristics of building structures are
sufficiently populated. However, data on dynamic characteristics of special structures such as bridges and
in particular cablestayed bridges are scarce.
The data bases are in general obtained from two sources:
− dynamic testing of structures (forced vibration, ambient vibration, etc),
− analyses of data from instrumented structures. This however is dependent on strongmotion response
data, which depends on a long duration project that is costly to implement and maintain.
1
ENEA – Casaccia Research Centre, Rome, Italy, paolo.clemente@casaccia.enea.it
2
U.S. Geological Survey, Menlo Park, CA, USA, celebi@usgs.gov
Whenever possible, lowamplitude tests are used to verify the results of analytical studies. Relevant
examples of dynamic testing were carried out on recent bridges, such as the Adamillo CableStayed
Bridge in Sevilla [1], the Pedestrian Bridge in Burlington, Vermont [2], the Tampico Bridge in Mexico [3]
and the Garigliano Bridge in Italy [4]. Wang & Huang [5] analysed the vibrations due to road roughness.
Fig. 1 Indiano CableStayed Bridge, Florence, Italy
The purpose of this paper is to present observed and analytical analyses to identify the dynamic
characteristics of Indiano Bridge in Florence, Italy. A general view of Indiano Bridge is shown in Figure 1.
The longitudinal section, the plan view, and the crosssection with the general dimensions of the bridge
are in Figure 2. The reasons for the choice of this particular bridge are:
− it is one of the longest spans in Italy,
− the structural system is simple and symmetric, thus analyses of the structural system can be performed
rather reliably for comparison with measurements,
− the main span is short so that during testing extensive cabling is not necessary,
− the towers are reasonably accessible for temporary deployment of sensors.
− the bridge has a “walkway” below the deck for pedestrians and bicyclists, which facilitates
deployment of sensors below the deck of the bridge rather than the top. Thus there is no interference
with the flow of the traffic. And finally,
− some of the characteristics of Indiano Bridge are common to several cable bridges. For example, the
presence of sliding supports at the ends causes oscillations of significant amplitude in the
longitudinal direction. The deteriorated conditions of the supports also cause additional damping.
These aspects make the analysis of the bridge quite complex but interesting.
DESCRIPTION OF THE STRUCTURE
The Indiano CableStayed Bridge over the Arno River in Florence was completed and opened to the traffic
in 1977 (Figure 1). The 189.1 m span girder of the bridge is simply supported by two piers, which are
structurally independent of the other parts of the structure. In the central portion of the girder (zone M in
Figure 2), whose length is 128.1 m, two boxes spaced of 6.0 m compose the crosssection. They are linked
one to another, at the upper and lower levels, by means of truss structures. As a result, the beam is
characterized by large torsional stiffness. The boxes have a width of 4.0 m and a height variable from 2.6
m to 1.6 m. Cantilever beams start from the boxes to support the external portion of the road. In the zones
near the ends (zones L and R in Figure 2) the crosssection becomes a three boxes crosssection.
The girder is suspended by six couples of fanshaped stays, starting at the tops of two steel towers. Cables
are regularly spread along the deck and are 3.0 m spaced from the centre line, therefore their contribution
in supporting the beam torsion is very low. Six cables constrained to an external gravity anchoring
compose each anchor cable.
Fig. 2 General dimensions of the bridge and sensor layout (see Table 1)
The pylons have steel box crosssection and a height of about 55.0 m from the ground. They are fully
constrained at their foundations, which are founded on large piles, and linked to the cable anchoring by
means of a prestressed concrete truss, which is supposed to support the horizontal component of the
tower stress. A footbridge is suspended to the girder. Due to its very high deformability, its contribution in
the dynamic behaviour of the bridge is considered to be negligible. The girder is supported at each pier by
means of four steel (knifeedge) roller bearings. At each bearing, lateral restraining plates stop the
transversal displacements. Four prestrained bars fixed into the pier and at the girder make the support
bilateral. At the time of the experimental tests they had been considerably damaged because of the
continued inversion of the reaction. As a result, the bearings were further stressed due to the uplift of the
girder so much so that their sliding function was seriously compromised. This condition is believed to
influence the damping characteristics of the bridge.
EXPERIMENTAL ANALYSIS
Eight Kinemetrics SS1 seismometers, an HP3566A signal conditioner and a laptop composed the
experimental setup. The synchronized signals from several seismometers were recorded by the
acquisition system and analysed in real time in order to have a first glance at the experimental data.
Seismometers were temporarily installed in different locations of the structure and the free field in 11
different configurations. The various configurations of deployment of these sensors are shown in Figure 2
and are summarized in Table 1, where for each location, the directions of the sensors are indicated (x =
longitudinal axis of the bridge, y = transversal axis, z = vertical axis). In configurations 9, 10 and 11, three
sensors were on the ground, about 200, 100 and 50 m from the bridge, respectively. Several timehistories
lasting 64 sec were recorded for each configuration to show repeatability of the vibrational characteristics
and to get average values of the characteristics. In the following, each record is identified by means of the
location and its direction (i.e., Ax is the sensor at A in the longitudinal direction x).
Tab. 1 Sensor locations and directions (see Figure 2)
Config. A B C D D′ E E′ F
Ground
(distance from C)
1 x – y x  y  z
2 X x – y x  y  z
3 x x  y x  y  z
4 x x  y x  y z z z
5 x x  y x  y y y y
6 x x  y x  y y y x
7 x x  y x  y z z z
8 x x  y x  y z z z
9 x  y  z x  y – z (200 m)
10 x  y  z x  y – z (100 m)
11 x  y  z x  y – z (50 m)
Data analysis
The dynamic characteristics of the structure from analyses of observed test data are provided in terms of
power spectral densities (PSD) and crossspectral densities (CSD). Spectra were estimated by using the
commercially available computer code MATLAB. Windowing was not used. In Figures 3 to 6 PSD and
CSD for selected tests 10, 16 and 22, are shown. In Table 3 the resonance frequencies with the relative
locations are listed. The modal shapes, obtained in the next numerical analysis, are shown in Figure 8.
The fundamental frequency, 0.61 Hz
3
, is clearly identifiable in the spectra of the records obtained at the
top of the tower in the longitudinal direction (Figure 3 – Bx) and in those of the deck, both in the vertical
(Figure 3 – Dz) and in the longitudinal direction (E'z). Ax and Bx being 180 degrees out of phase (Figure
5) and Dz, Ez and Fz in phase, the first modal shape happens to be symmetric around the midspan.
Clearly, Bx and Ez are in phase, inferring that when the towers move outwards, the deck moves upwards
[Figure 11 – Mode 1].
The resonance frequency 0.81 Hz is very apparent in the spectrum of records on the deck in the
longitudinal direction (E'z) and the crossspectra with the records of the top of the towers confirm that
3
Just after the completion of the construction, the bridge was tested [6]. The structure was excited by
dropping a suspended mass of 0.8 MNs
2
/m from the middle of the span. In this way only the first mode
was excited. An associated frequency of 0.61 Hz was estimated.
(Figure 6 –Bx–E'z). The modal shape associated to this frequency consists in the longitudinal movement of
tower and deck (Figure 11 – Mode 2).
Peaks at frequency 0.87 Hz identified in the spectra of the deck records in the transverse direction (Dy, Ey
and Fy) indicate the fundamental transverse deck frequency. The corresponding modal shape (Figure 11 –
Mode 3) represents the first mode in the transverse direction for the girder. The towers do not contribute
to this mode.
Peaks at frequencies 1.04 and 1.15 Hz are apparent in the spectra of records at the tower top in the
transverse direction (Figure 3 – By). The deck does not contribute to this mode. Records of sensors Ez and
E'z, in the vertical direction during Test 25, are 180 degree out of phase at 1.15 Hz.
Peaks at frequency 1.41 Hz are apparent in the spectra of records at the tower top in the longitudinal
direction (Figure 3 – Bx). Low peaks are also visible in the spectra of the vertical records on the girder
(Figure 3 – Dz).
Tab. 3 Experimental resonance frequencies and estimated damping
Sensor location Freq. (Hz)
Tower longitudinal 0.61 1.41 1.61 2.25
Tower transversal 1.04 1.15 2.61
Deck vertical 0.61 1.04 1.15 1.41 1.61 2.25
Deck. transversal 0.87 2.61
Deck longitudinal 0.61 0.81 1.41 1.61
Damping 6.5% 6% 4.5% 4.8%   3% 2.5%
Fig. 3 Power Spectral Densities (PSD). Sensors Bx, By, Dz, Cy
At frequency 1.61 Hz, Ax and Bx (Figure 4) are 180 degrees out of phase, so they move antisymmetrically
around the midspan. Dz and Fz are in phase at this frequency, while Ez and Fz are 180 degrees out of
phase, as a result the girder shaped symmetrically with three half waves (Figure 11 – Mode 7).
At frequency 2.25 Hz Ax and Bx are in phase, Dz and Fz out of phase. A twowave mode shape is
associated with this frequency. In this case also the towers take part in the mode (Figure 11 – Mode 8).
Peaks at frequency 2.61 (Dy, Ey and Fy) are relative to the second transverse mode of the girder (Figure
11 – Mode 9).
Fig. 4 Cross Spectral Densities (CSD), Phase Factor and Coherence Function (dotted line)
Sensors AxBx, BxEz
Fig. 5 CSD Sensors DyEy Fig. 6 CSD Sensors Bx E'x
The critical viscous damping ratios were estimated by the half power bandwidth method and are
summarized in Table 3. The values of damping obtained are very high compared with the usual ones for
cable bridges. This is due to the presence of several sources of damping in joints and especially in
bearings. It is restated herein that damping due to sliding supports increases for oscillation with very small
amplitude [7, 8].
Consideration about site conditions
A series of noise measurements at the free field around the bridge on the north side of the river were made
to assess possible effect of the site conditions on the response and behaviour of the. We kept three
seismometers at the basement of one tower (location C) and moved the other three in three different
positions, about 50, 100 and 200 m away from the same tower (B). For each configuration, four tests
lasting 64 s were carried out. Recorded data were analysed in the frequency domain [0, 5 Hz]. Some of
the natural frequencies found for the bridge appear in the ratios of PSD of horizontal to vertical
components [9] at each position only for a few of the records. This is probably due to the different
excitation level from the bridge traffic. A common feature for these ratios is the presence of two peaks at
frequencies below 0.5 Hz.
The ratios between the PSDs of the same components at different position confirm the presence of the two
peaks below 0.5 Hz. Since we are in an alluvial site these frequencies are supposed to be representative of
the soil behaviour. Further investigations of the records have been excluded because this frequency value
is at the limit range of the sensor response.
From this analysis we conclude that there is not any apparent relationship between the behaviour of the
soil and the natural frequencies of the bridge.
MATHEMATICAL MODEL
A finite element model of the bridge was developed using the geometrical and mechanical characteristics
of the bridge from the structural design drawings obtained via the Technical Office of the City Hall in
Florence. The finite element computer code COSMOS (Ver. 2.5) was used for this purpose.
1
2
3
4
x
z
y
Fig. 7 Finite element model
Assumptions
The primary assumption in the finite element analysis performed is that the structure is assumed to be
linear elastic. The towers are assumed to be fully constrained at their basements. The girder is simply
supported at its ends; vertical displacements and torsional rotation are not allowed.
The girder and the pylons have been modelled by using 3Dbeam elements. Rigid limbs have been offset
both from the girder and the towers, in order to accommodate the cable attachments (Figure 7). The
bending and torsional stiffness of the girder are variable along the span. In the finite element model the
girder has been divided in three portions, in each of them the geometrical characteristics have been
assumed to be constant and equal to the average values. These are listed in Table 4, in which A is the cross
sectional area, I
y
and I
z
the second moments, relative to the transversal axes of the crosssection
(horizontal and vertical, respectively), and J is the torsion factor. The geometrical characteristics of the
towers change along their axis. These geometrical characteristics, based on design drawings as
summarized in Table 5, are used to develop the mathematical model.
Tab. 4 Geometrical properties and Young’s modulus of the girder beam elements
∆x A (m
2
) I
y
(m
4
) I
z
(m
4
) J (m
4
) E (MPa)
0 < x < 30.5 m 0.8637 0.7554 24.29 1.825 210000
30.5 < x < 168.6 m 0.8427 0.6753 24.29 1.136 210000
168.6 < x < 189.1 m 0.8637 0.7554 24.29 1.825 210000
Tab. 5 Geometrical properties and Young’s modulus of the tower beam elements
∆z A (m
2
) I
y
(m
4
) I
x
(m
4
) J (m
4
) E (MPa)
0 < z < 12.7 m 0.4782 0.4453 1.075 0.8872 210000
12.7 < z < 25.5 m 0.4208 0.3795 0.6373 0.6771 210000
25.5 < z < 35.5 m 0.3549 0.3050 0.3557 0.4790 210000
35.5 < z < 48.3 m 0.5632 0.3951 0.4615 0.6582 210000
Cables have been assumed as linear elastic truss elements. The reduction of their axial stiffness due to the
sag has been accounted for by assuming reduced Young’s modules, relative to their configuration under
dead loads, given by the relation ship:
( )
¸
¸
.
!
\

+
+ + =
24
2
24
1
1
2
2 *
ϕ σ
ϕ
E
E E
Tab. 6 Cross sectional areas and apparent moduli of the cables
Cable Number of
ropes
A (cm
2
) (for
each rope)
E (MPa)
1 2 * 3 119.7 200800
2 2 * 1 134.7 203300
3 2 * 1 140.5 200900
4 2 * 2 101.0 163800
where E is the Young’s modulus of the cable material, σ is the stress under dead loads and ϕ = γλ/σ is a
shape parameter, γ and λ being the weight per unit volume and the cable span, respectively [10, 11]. The
crosssectional areas and the elastic modules of the cables are provided in Table 6.
The mass of the girder (0.140 MN s
2
/m
2
) has been increased by 10% to account for the vehicles and
additional masses during the tests. The resulting distributed mass of 0.154 MN s
2
/m
2
and the mass inertia
around xaxis of 4.02 MN s
2
/m have been lumped at appropriate girder model nodes.
Modal Analysis
The modal analysis resulted in the frequencies listed in Table 7. The first nine modal shapes are plotted in
Figure 8. The modal shapes of the numerical model and the corresponding frequencies are quite similar to
the experimental ones. Discrepancies have been found only in the first resonance frequencies
corresponding to the transverse and longitudinal motions of the deck. However these discrepancies can be
endorsed to the differences between the actual distribution of the stiffness and the assumed ones and to
the presence of unknown masses at the moment of the tests. The nonlinear behaviour of the supports due
to friction played an important role.
Tab. 7 Frequencies of the numerical model and experimental resonance
Mode
number
Numerical
Frequency
(Hz)
Dominant displacements Experimental
Frequency
(Hz)
1 0.61 Deck Vertical 0.61
2 0.64 Deck Longitudinal 0.81
3 0.77 Deck Transverse 0.87
4 1.12 Deck Torsion & Tower Trans. 1.04
5 1.13 Deck Torsion & Tower Trans. 1.15
6 1.37 Tower Longitudinal 1.41
7 1.62 Deck Vertical 1.61
8 2.41 Deck Vertical 2.25
9 3.09 Deck Transverse 2.61
SEISMIC RESPONSE ANALYSIS OF THE BRIDGE
Seismic Hazard Analysis
Molin and Paciello [12], using both a probabilistic approach and the analysis of a local macroseismic
catalogue, assessed the seismic hazard in Florence. Only the main points (relevant to the following
spectrum response analysis) of their probabilistic approach are reported herein.
The probabilistic hazard assessment was based on the wellknown Cornell’s methodology [13], which
models earthquake occurrences in a given time interval according to a Poisson distribution. Seismic data
from GNDT (Italian National Group for Protection against Earthquakes) catalogue were used, in which
only events with M
s
≥ 4.0 or with macroseismic intensity greater than V MCS (MercalliCancaniSieberg
[14]), are reported, and the dependent events (aftershocks and foreshocks) are removed. The seismic
sources of interest for the site were selected from the seismotectonic zonation of Italian territory proposed
by Scandone et al. [15]. The attenuation laws proposed by Sabetta and Pugliese [16, 17] were used for
ground motion parameters.
The analysis was performed in terms of macroseismic intensity, peak ground acceleration, peak ground
velocity and pseudovelocity response spectra. The main results can be summarized as follows. The
maximum intensity historically observed in Florence is VII MCS. Therefore, MCS VIII intensity is
considered as the maximum credible for the city. Thus, from these the peak ground acceleration values
with 90% nonexceedence probability, in 50 and 500 years respectively, are determined as 145 and 219
cm/s
2
for shallow alluvium, and 95 and 157 cm/s
2
for deep alluvium sites. Since, the Indiano Bridge site is
a deep alluvial site, in Figure 9 the uniform response spectra with 5% damping obtained for deep alluvium
sites that shows the pseudovelocity values having 90% nonexceedence probability for 50, 100 and 200
years respectively are provided. The highest pseudovelocity values are between 1 and 2 Hz.
Fig. 8 Modal shapes
Response spectrum analysis
In the response spectrum analysis of the bridge, the spectrum for 5% damping and 200 years return period
is used in the three main axis of the bridge, separately. The directional modal effects are considered by
square root of the sum of squares of the response corresponding to directional input. Because of the
structural symmetry the effects due to the input in the different directions were almost independent one of
the others.
The results of the response spectrum analysis have been compared with those of the static analysis, carried
out assuming the following load conditions, suggested by the Italian Code for design, construction and test
of bridges [18]:
a) four lanes of vehicles, two of 30.0 KN/m and two of 15.0 KN/m, symmetrically deployed with
reference to the longitudinal axis of the bridge;
b) a longitudinal load whose resultant is equal to 3.0 KN/m, simulating a braking column of vehicles;
c) a transversal loads equal to 2.5 KN/m
2
acting on the bridge in presence of a vehicle column of 3.00 m
height, simulating the wind excitation. Obviously a correct analysis of wind effects would require a
more detailed study to account for the dynamic effect of the wind.
The following displacement parameters have been considered:
• maximum vertical displacement of the deck u
z
, and the longitudinal displacement of the tower u
tx
to
compare with that due to load a;
• maximum longitudinal oscillation of the deck u
x
, to compare with that due to load b;
• maximum transversal displacement of the beam u
y
and the tower u
ty
, to compare with those due to the
wind excitation (load c).
and the following force parameters:
• maximum bending moment M
y
in the beam (load a),
• maximum bending moment M
z
in the beam (load c),
• maximum force in the anchor stay S
o
(load a),
• maximum force in the longest stay S
m
(load a),
• maximum axial force in the tower N
p
(load a).
Tab. 8 Comparison between static and dynamic analysis
Parameter Response
spectrum analysis
Static analysis
(live loads)
u
z
(m) 0.100 0.469
u
x
(m) 0.040 0.006
u
y
(m) 0.090 0.048
u
ty
(m) 0.074 0.026
u
tx
(m) 0.025 0.100
M
y
(MNm) 112 246
M
z
(MNm) 1292 657
S
o
(MN) (for each
rope)
6.1 24.4
S
m
(MN) 5.9 19.9
N
p
(MN) 40.4 158
The results obtained are summarized in Table 8. The effects of the hypothesised earthquake are greater
than those of the static analysis, especially in terms of horizontal displacements. In more detail:
• the vertical displacement of the deck and the longitudinal displacement of the tower due to static
loads are lower than those due to seismic actions;
• the transversal displacement is much higher than the corresponding one due to the wind load;
• the longitudinal displacement is much higher than that due to the longitudinal force generated by a
braking vehicle column;
• the value of the bending moment M
y
is not negligible if compared to the moment due to static loads,
while the bending moment M
z
is about twice that one due to wind action;
• the axial forces in the tower and in the cables are much lower than the corresponding one due to static
loadings.
CONCLUSIONS
The dynamic response assessment and consequential seismic analysis of the Indiano CableStayed Bridge
was performed by means of (a) observed test data and subsequent spectral analyses and (b) analytical
models using design geometric parameters . Contemporarily, the study of the seismic hazard in Florence
site produced the uniform response spectra in terms of pseudovelocities. These were used as input in the
dynamic analysis of the bridge, carried out in terms of response spectrum analysis.
The results can be summarized as follows:
• Several resonance frequencies are clearly identified from the observed dynamic response data and
also from modal analyses using a mathematical model. The modal shapes inferred from observed data
are very similar to those obtained from modal analyses using the mathematical model;
• The structure shows a modal shape whose prevalent displacements are in the longitudinal direction
due to motion of the towers;
• Trafficinduced vibration amplitudes, both in vertical and in longitudinal directions, are very high, so
that to determine frequent needs of maintenance works;
• Values of estimated structural damping are very high compared with those determined for cable
bridges. This occurrence is related to the presence of several sources of damping (friction) in joints
and in bearings;
• Even thought Florence is not a high intensity seismic site, the expected earthquake based on MCS
intensity VIII would cause significant increase of stresses in the bridge.
One of the characteristics of the dynamic behaviour of the Indiano Bridge is the presence of longitudinal
vibrations of high amplitude. The absence of longitudinal restraints is usual in cablestayed bridges, but in
this case the above occurrence is amplified because of the absence of side cablestayed spans and because
the system is not selfanchored. From the dynamic point of view, one of the end bearings could be a fixed
hinge, but the static consequences of such a kind of change should be analysed in detail (changes in the
cable stresses under nonsymmetrical loads, stresses in the piers due to horizontal forces).
ACNOWLEDGEMENTS
Funding by NATO (Research Grant No. CRG 960109) is gratefully acknowledged. The authors wish also
to acknowledge the Technical Office of the Florence City Hall for having allowed this study.
REFERENCES
1. Casas J. R. “Fullscale dynamic testing of the Adamillo cablestayed bridge in Sevilla (Spain)”, Int J.
Earthquake Engineering and Structural Dynamics, John Wiley & Sons, 1995; 24(1): 3551.
2. GardnerMorse M.G. & Huston D.R. “Modal Identification of CableStayed Pedestrian Bridge”,
Journal of Structural Engineering, ASCE, 1993; 119(11): 33843404.
3. MuriàVila D., Gomez R & King C. “Dynamic Structural Properties of CableStayed Tampico
Bridge”, Journal of Structural Engineering, ASCE, 1991; 117(11): 33963416.
4. Clemente P., Marulo S., Lecce L. & Bifulco A. “Experimental modal analysis of the Garigliano cable
stayed bridge”, Int. J. Soil Dynamics and Earthquake Engineering, Elsevier Science, 1998; 17(78):
485493.
5. Wang T.L. & Huang D. “CableStayed Bridge Vibration due to Road Surface Roughness”, Journal of
Structural Engineering, ASCE, 1992; 118(5): 13541374.
6. Augusti G., Chiarugi A., Vignoli A. “Analisi sperimentale di un ponte strallato”, Proc. 7th Italian
Conference on Steel Structures (Torino 1979), C.T.A. Milano, 1979 (in Italian).
7. Kawashima K. & Unjoh S. “Damping Characteristics of CableStayed Bridges Associated with
Energy Dissipation at Movable Bearings”, Proc., Japan Society of Structural Engineering and
Earthquake Engineering, 1989; No. 404, 6(1): 145152.
8. Kawashima K., Unjoh S. & Tunomoto M. “Estimation of Damping Ratio of CableStayed Bridges for
Seismic Design”, Journal of Structural Engineering, ASCE, 1993; 119(4): 10151031.
9. Nakamura Y., “A method for Dynamic Characteristics estimation of Subsurface using Microtremor
on the Ground Surface”, QR of RTRI, 1989; 30(1): 2533.
10. Clemente P., D”Apuzzo M. “Analisi del modello generalizzato di ponte strallato” Fondazione
Politecnica per il Mezzogiorno d”Italia, Giannini Napoli, 1990; No. 162 (in Italian).
11. D’Apuzzo M., Clemente P. “Progetto preliminare dei ponti strallati”, Autostrade, AutostradeItalstat
Roma, 1990; 3: 7078 (in Italian).
12. Molin D. & Paciello A. “Seismic Hazard Assessment in Florence City Italy”, Journal of Earthquake
Engineering, Imperial College Press, 1999; 3(4): 475494.
13. Cornell C. A. “Engineering seismic risk analysis”, Bull. Seism Soc. Am., 1968; 58(5): 15831606.
14. “MCS, MercalliCancaniSieberg macroseismic scale”, From Sieberg A., Geologie der Erdbeben,
Handbuch der Geophysic, Tab. 2 and 3, Berlin, 1930.
15. Scandone P., Patacca E., Meletti C., Bellatalla M., Perilli N. & Santini U. “Struttura geologica,
evoluzione cinematica e schema sismotettonico della penisola italiana”, Atti Convegno GNDT (Pisa,
1990), GNDT, 1991; 1.
16. Sabetta F. & Pugliese A. “Attenuation of peak horizontal acceleration and velocity from Italian
strongmotion records”, Bull. Seism. Soc. Am., 1987; 77(5): 14911513.
17. Pugliese A. & Sabetta F. “Stima degli spettri di risposta da registrazioni di forti terremoti italiani”,
Ingegneria Sismica, 1989; 2: 314 (in Italian).
18. Ministero Lavori Pubblici “Aggiornamento delle norme tecniche per la progettazione, l’esecuzione e
il collaudo dei ponti stradali  D.M. LL.PP. 4 maggio1990”, G.U. Repubblica Italiana, 29 Jan. 1991.
Relevant examples of dynamic testing were carried out on recent bridges. Wang & Huang [5] analysed the vibrations due to road roughness. − the towers are reasonably accessible for temporary deployment of sensors. In the central portion of the girder (zone M in Figure 2). The 189. For example. whose length is 128. 1 Indiano CableStayed Bridge. Thus there is no interference with the flow of the traffic. − some of the characteristics of Indiano Bridge are common to several cable bridges.Whenever possible. Florence. Italy The purpose of this paper is to present observed and analytical analyses to identify the dynamic characteristics of Indiano Bridge in Florence. the Tampico Bridge in Mexico [3] and the Garigliano Bridge in Italy [4]. such as the Adamillo CableStayed Bridge in Sevilla [1]. which facilitates deployment of sensors below the deck of the bridge rather than the top. These aspects make the analysis of the bridge quite complex but interesting. Vermont [2]. Fig. − the bridge has a “walkway” below the deck for pedestrians and bicyclists. the Pedestrian Bridge in Burlington. two boxes spaced of 6. The deteriorated conditions of the supports also cause additional damping. the plan view. Italy. and the crosssection with the general dimensions of the bridge are in Figure 2.0 m compose the crosssection. thus analyses of the structural system can be performed rather reliably for comparison with measurements. The longitudinal section. A general view of Indiano Bridge is shown in Figure 1. DESCRIPTION OF THE STRUCTURE The Indiano CableStayed Bridge over the Arno River in Florence was completed and opened to the traffic in 1977 (Figure 1). And finally.1 m. − the main span is short so that during testing extensive cabling is not necessary. the presence of sliding supports at the ends causes oscillations of significant amplitude in the longitudinal direction. which are structurally independent of the other parts of the structure. − the structural system is simple and symmetric. They are linked . The reasons for the choice of this particular bridge are: − it is one of the longest spans in Italy.1 m span girder of the bridge is simply supported by two piers. lowamplitude tests are used to verify the results of analytical studies.
and linked to the cable anchoring by means of a prestressed concrete truss. Six cables constrained to an external gravity anchoring compose each anchor cable.6 m. . an HP3566A signal conditioner and a laptop composed the experimental setup. the beam is characterized by large torsional stiffness. starting at the tops of two steel towers. the bearings were further stressed due to the uplift of the girder so much so that their sliding function was seriously compromised.0 m from the ground. At each bearing. which is supposed to support the horizontal component of the tower stress. At the time of the experimental tests they had been considerably damaged because of the continued inversion of the reaction. Four prestrained bars fixed into the pier and at the girder make the support bilateral.0 m and a height variable from 2. The girder is supported at each pier by means of four steel (knifeedge) roller bearings.0 m spaced from the centre line. This condition is believed to influence the damping characteristics of the bridge. therefore their contribution in supporting the beam torsion is very low. by means of truss structures. In the zones near the ends (zones L and R in Figure 2) the crosssection becomes a three boxes crosssection. Cables are regularly spread along the deck and are 3. Due to its very high deformability. Cantilever beams start from the boxes to support the external portion of the road. EXPERIMENTAL ANALYSIS Eight Kinemetrics SS1 seismometers. at the upper and lower levels. 2 General dimensions of the bridge and sensor layout (see Table 1) The pylons have steel box crosssection and a height of about 55. Fig. lateral restraining plates stop the transversal displacements. The girder is suspended by six couples of fanshaped stays. The synchronized signals from several seismometers were recorded by the acquisition system and analysed in real time in order to have a first glance at the experimental data. As a result.one to another. As a result. They are fully constrained at their foundations. A footbridge is suspended to the girder. which are founded on large piles. The boxes have a width of 4.6 m to 1. its contribution in the dynamic behaviour of the bridge is considered to be negligible.
obtained in the next numerical analysis. The various configurations of deployment of these sensors are shown in Figure 2 and are summarized in Table 1. 3 . Spectra were estimated by using the commercially available computer code MATLAB.Seismometers were temporarily installed in different locations of the structure and the free field in 11 different configurations. In Figures 3 to 6 PSD and CSD for selected tests 10. Windowing was not used. are shown. y = transversal axis. Ax is the sensor at A in the longitudinal direction x). 1 Sensor locations and directions (see Figure 2) Config. The fundamental frequency.y – z (200 m) x . The structure was excited by dropping a suspended mass of 0. the first modal shape happens to be symmetric around the midspan. respectively. the directions of the sensors are indicated (x = longitudinal axis of the bridge. Ez and Fz in phase. Bx and Ez are in phase. Tab. 0. where for each location. In the following. 100 and 50 m from the bridge. are shown in Figure 8. z = vertical axis). both in the vertical (Figure 3 – Dz) and in the longitudinal direction (E'z). The resonance frequency 0. The modal shapes. three sensors were on the ground. In this way only the first mode was excited.61 Hz3. Clearly.8 MNs2/m from the middle of the span. An associated frequency of 0. In configurations 9. the deck moves upwards [Figure 11 – Mode 1]. In Table 3 the resonance frequencies with the relative locations are listed.y – z (50 m) Data analysis The dynamic characteristics of the structure from analyses of observed test data are provided in terms of power spectral densities (PSD) and crossspectral densities (CSD).e. each record is identified by means of the location and its direction (i.61 Hz was estimated.. 10 and 11. inferring that when the towers move outwards. Ax and Bx being 180 degrees out of phase (Figure 5) and Dz. Several timehistories lasting 64 sec were recorded for each configuration to show repeatability of the vibrational characteristics and to get average values of the characteristics. is clearly identifiable in the spectra of the records obtained at the top of the tower in the longitudinal direction (Figure 3 – Bx) and in those of the deck. the bridge was tested [6]. 16 and 22. 1 2 3 4 5 6 7 8 9 10 11 A B x–y C xyz xyz xyz xy xy xy xy xy xyz xyz xyz D D′ E E′ F Ground (distance from C) X x x x x x x x–y xy xy xy xy xy xy z y y z z z z y y z z x z z y x .81 Hz is very apparent in the spectrum of records on the deck in the longitudinal direction (E'z) and the crossspectra with the records of the top of the towers confirm that Just after the completion of the construction. about 200.y – z (100 m) x .
04 1. Peaks at frequency 0.61 Damping 6.41 Hz are apparent in the spectra of records at the tower top in the longitudinal direction (Figure 3 – Bx).15 1.61 Freq. Cy . The modal shape associated to this frequency consists in the longitudinal movement of tower and deck (Figure 11 – Mode 2). are 180 degree out of phase at 1.61 1. Ey and Fy) indicate the fundamental transverse deck frequency. Sensors Bx. Peaks at frequency 1.(Figure 6 –Bx–E'z).15 1. The towers do not contribute to this mode.41 1. Dz.87 1.61 2. The corresponding modal shape (Figure 11 – Mode 3) represents the first mode in the transverse direction for the girder.25 2. 3 Power Spectral Densities (PSD). By.15 Hz.5% 6% 4.61 0. transversal Deck longitudinal 0.8%   3% 2.04 and 1.61 2. Peaks at frequencies 1.41 1.81 0.41 1. 3 Experimental resonance frequencies and estimated damping Sensor location Tower longitudinal Tower transversal Deck vertical Deck. Records of sensors Ez and E'z.87 Hz identified in the spectra of the deck records in the transverse direction (Dy.61 0.5% Fig. (Hz) 1. Low peaks are also visible in the spectra of the vertical records on the girder (Figure 3 – Dz).15 Hz are apparent in the spectra of records at the tower top in the transverse direction (Figure 3 – By).61 0.04 1. The deck does not contribute to this mode. in the vertical direction during Test 25.5% 4. Tab.25 2.
At frequency 2. 4 Cross Spectral Densities (CSD). Ey and Fy) are relative to the second transverse mode of the girder (Figure 11 – Mode 9).61 Hz. Dz and Fz are in phase at this frequency. so they move antisymmetrically around the midspan. Peaks at frequency 2. 6 CSD Sensors Bx. Dz and Fz out of phase.61 (Dy. A twowave mode shape is associated with this frequency.25 Hz Ax and Bx are in phase.E'x . In this case also the towers take part in the mode (Figure 11 – Mode 8).At frequency 1. BxEz Fig. Ax and Bx (Figure 4) are 180 degrees out of phase. as a result the girder shaped symmetrically with three half waves (Figure 11 – Mode 7). Fig. 5 CSD Sensors DyEy Fig. Phase Factor and Coherence Function (dotted line) Sensors AxBx. while Ez and Fz are 180 degrees out of phase.
The values of damping obtained are very high compared with the usual ones for cable bridges. Since we are in an alluvial site these frequencies are supposed to be representative of the soil behaviour. Further investigations of the records have been excluded because this frequency value is at the limit range of the sensor response. about 50. 4 2 1 z y x 3 Fig. 100 and 200 m away from the same tower (B). We kept three seismometers at the basement of one tower (location C) and moved the other three in three different positions. 5 Hz]. From this analysis we conclude that there is not any apparent relationship between the behaviour of the soil and the natural frequencies of the bridge. This is due to the presence of several sources of damping in joints and especially in bearings.5 Hz.The critical viscous damping ratios were estimated by the half power bandwidth method and are summarized in Table 3. This is probably due to the different excitation level from the bridge traffic. Recorded data were analysed in the frequency domain [0. It is restated herein that damping due to sliding supports increases for oscillation with very small amplitude [7. The ratios between the PSDs of the same components at different position confirm the presence of the two peaks below 0. The finite element computer code COSMOS (Ver. For each configuration. Some of the natural frequencies found for the bridge appear in the ratios of PSD of horizontal to vertical components [9] at each position only for a few of the records. A common feature for these ratios is the presence of two peaks at frequencies below 0.5) was used for this purpose. 2. four tests lasting 64 s were carried out. MATHEMATICAL MODEL A finite element model of the bridge was developed using the geometrical and mechanical characteristics of the bridge from the structural design drawings obtained via the Technical Office of the City Hall in Florence. Consideration about site conditions A series of noise measurements at the free field around the bridge on the north side of the river were made to assess possible effect of the site conditions on the response and behaviour of the. 7 Finite element model .5 Hz. 8].
5 m 35.825 1. The bending and torsional stiffness of the girder are variable along the span.4615 J (m4) 0. In the finite element model the girder has been divided in three portions. and J is the torsion factor. The towers are assumed to be fully constrained at their basements. The girder is simply supported at its ends. Iy and Iz the second moments. The girder and the pylons have been modelled by using 3Dbeam elements.5 < x < 168. based on design drawings as summarized in Table 5. 5 Geometrical properties and Young’s modulus of the tower beam elements ∆z 0 < z < 12. given by the relation ship: 1 2E E * = E 1 + ϕ 2 + 24 σ ϕ 2 + 24 ( ) Tab.4790 0.8872 0.7 140.Assumptions The primary assumption in the finite element analysis performed is that the structure is assumed to be linear elastic. in order to accommodate the cable attachments (Figure 7).6 < x < 189.4453 0. in which A is the cross sectional area.5 101. 4 Geometrical properties and Young’s modulus of the girder beam elements ∆x 0 < x < 30.29 24.5 m 30.29 24.29 J (m4) 1.825 E (MPa) 210000 210000 210000 Tab.075 0.5 m 25.3795 0. relative to their configuration under dead loads.8427 0. in each of them the geometrical characteristics have been assumed to be constant and equal to the average values.4782 0.3 m A (m2) 0.7554 Iz (m4) 24.4208 0.6582 E (MPa) 210000 210000 210000 210000 Cables have been assumed as linear elastic truss elements.7 m 12.6753 0.0 E (MPa) 200800 203300 200900 163800 .6 m 168. These are listed in Table 4. The geometrical characteristics of the towers change along their axis.5632 Iy (m4) 0.6771 0.7554 0. vertical displacements and torsional rotation are not allowed. respectively).136 1. These geometrical characteristics. are used to develop the mathematical model.3050 0.5 < z < 35. relative to the transversal axes of the crosssection (horizontal and vertical.5 < z < 48. 6 Cross sectional areas and apparent moduli of the cables Cable 1 2 3 4 Number of ropes 2*3 2*1 2*1 2*2 A (cm2) (for each rope) 119. Rigid limbs have been offset both from the girder and the towers.8637 0.8637 Iy (m4) 0. The reduction of their axial stiffness due to the sag has been accounted for by assuming reduced Young’s modules.3549 0.3951 Ix (m4) 1.7 134. Tab.7 < z < 25.6373 0.3557 0.1 m A (m2) 0.
where E is the Young’s modulus of the cable material. assessed the seismic hazard in Florence.13 1.37 1.0 or with macroseismic intensity greater than V MCS (MercalliCancaniSieberg [14]).81 0. Discrepancies have been found only in the first resonance frequencies corresponding to the transverse and longitudinal motions of the deck. Only the main points (relevant to the following spectrum response analysis) of their probabilistic approach are reported herein.64 0. Tab. The first nine modal shapes are plotted in Figure 8. 7 Frequencies of the numerical model and experimental resonance Mode number 1 2 3 4 5 6 7 8 9 Numerical Frequency (Hz) 0. However these discrepancies can be endorsed to the differences between the actual distribution of the stiffness and the assumed ones and to the presence of unknown masses at the moment of the tests.12 1. The nonlinear behaviour of the supports due to friction played an important role. in which only events with Ms ≥ 4. The crosssectional areas and the elastic modules of the cables are provided in Table 6.61 0. are reported.62 2.154 MN s2/m2 and the mass inertia around xaxis of 4.140 MN s2/m2) has been increased by 10% to account for the vehicles and additional masses during the tests.15 1.61 2.09 Dominant displacements Experimental Frequency (Hz) 0. Seismic data from GNDT (Italian National Group for Protection against Earthquakes) catalogue were used.77 1. σ is the stress under dead loads and ϕ = γλ/σ is a shape parameter. Modal Analysis The modal analysis resulted in the frequencies listed in Table 7.41 1.25 2. The probabilistic hazard assessment was based on the wellknown Cornell’s methodology [13]. Tower Longitudinal Deck Vertical Deck Vertical Deck Transverse SEISMIC RESPONSE ANALYSIS OF THE BRIDGE Seismic Hazard Analysis Molin and Paciello [12]. using both a probabilistic approach and the analysis of a local macroseismic catalogue.61 Deck Vertical Deck Longitudinal Deck Transverse Deck Torsion & Tower Trans. which models earthquake occurrences in a given time interval according to a Poisson distribution.04 1. γ and λ being the weight per unit volume and the cable span. 11]. and the dependent events (aftershocks and foreshocks) are removed. The seismic .41 3. respectively [10. The modal shapes of the numerical model and the corresponding frequencies are quite similar to the experimental ones. The mass of the girder (0.61 0.87 1.02 MN s2/m have been lumped at appropriate girder model nodes. Deck Torsion & Tower Trans. The resulting distributed mass of 0.
[15]. Thus. The attenuation laws proposed by Sabetta and Pugliese [16. The analysis was performed in terms of macroseismic intensity. Since. separately. 100 and 200 years respectively are provided. the spectrum for 5% damping and 200 years return period is used in the three main axis of the bridge. in Figure 9 the uniform response spectra with 5% damping obtained for deep alluvium sites that shows the pseudovelocity values having 90% nonexceedence probability for 50. 8 Modal shapes Response spectrum analysis In the response spectrum analysis of the bridge. The main results can be summarized as follows. and 95 and 157 cm/s2 for deep alluvium sites. peak ground acceleration. are determined as 145 and 219 cm/s2 for shallow alluvium. The directional modal effects are considered by square root of the sum of squares of the response corresponding to directional input. The highest pseudovelocity values are between 1 and 2 Hz. Fig. 17] were used for ground motion parameters. Therefore. . the Indiano Bridge site is a deep alluvial site. from these the peak ground acceleration values with 90% nonexceedence probability.sources of interest for the site were selected from the seismotectonic zonation of Italian territory proposed by Scandone et al. MCS VIII intensity is considered as the maximum credible for the city. The maximum intensity historically observed in Florence is VII MCS. in 50 and 500 years respectively. Because of the structural symmetry the effects due to the input in the different directions were almost independent one of the others. peak ground velocity and pseudovelocity response spectra.
026 0.1 5.048 0.100 0. to compare with those due to the wind excitation (load c).00 m height. • the transversal displacement is much higher than the corresponding one due to the wind load. Obviously a correct analysis of wind effects would require a more detailed study to account for the dynamic effect of the wind.025 112 1292 6.090 0. symmetrically deployed with reference to the longitudinal axis of the bridge. 8 Comparison between static and dynamic analysis Parameter uz (m) ux (m) Response spectrum analysis 0.006 uy (m) uty (m) utx (m) My (MNm) Mz (MNm) So (MN) (for each rope) Sm (MN) Np (MN) 0.4 19.0 KN/m. .074 0. • maximum longitudinal oscillation of the deck ux. and the following force parameters: • maximum bending moment My in the beam (load a). to compare with that due to load b.4 0. • maximum axial force in the tower Np (load a).9 40.100 246 657 24. simulating the wind excitation. b) a longitudinal load whose resultant is equal to 3. and the longitudinal displacement of the tower utx to compare with that due to load a.0 KN/m. construction and test of bridges [18]: a) four lanes of vehicles. The following displacement parameters have been considered: • maximum vertical displacement of the deck uz. carried out assuming the following load conditions. • the longitudinal displacement is much higher than that due to the longitudinal force generated by a braking vehicle column.040 Static analysis (live loads) 0.0 KN/m and two of 15. two of 30. suggested by the Italian Code for design. In more detail: • the vertical displacement of the deck and the longitudinal displacement of the tower due to static loads are lower than those due to seismic actions. • maximum force in the anchor stay So (load a). • maximum force in the longest stay Sm (load a).469 0.5 KN/m2 acting on the bridge in presence of a vehicle column of 3. c) a transversal loads equal to 2.The results of the response spectrum analysis have been compared with those of the static analysis. simulating a braking column of vehicles. Tab. The effects of the hypothesised earthquake are greater than those of the static analysis. • maximum bending moment Mz in the beam (load c). • maximum transversal displacement of the beam uy and the tower uty.9 158 The results obtained are summarized in Table 8. especially in terms of horizontal displacements.
• Values of estimated structural damping are very high compared with those determined for cable bridges. From the dynamic point of view. Contemporarily.. This occurrence is related to the presence of several sources of damping (friction) in joints and in bearings. 24(1): 3551. 1998. 2. The results can be summarized as follows: • Several resonance frequencies are clearly identified from the observed dynamic response data and also from modal analyses using a mathematical model. The absence of longitudinal restraints is usual in cablestayed bridges. CONCLUSIONS The dynamic response assessment and consequential seismic analysis of the Indiano CableStayed Bridge was performed by means of (a) observed test data and subsequent spectral analyses and (b) analytical models using design geometric parameters . . Soil Dynamics and Earthquake Engineering. • Trafficinduced vibration amplitudes. John Wiley & Sons.. but the static consequences of such a kind of change should be analysed in detail (changes in the cable stresses under nonsymmetrical loads. “Fullscale dynamic testing of the Adamillo cablestayed bridge in Sevilla (Spain)”. Casas J. • The structure shows a modal shape whose prevalent displacements are in the longitudinal direction due to motion of the towers. Int J. the expected earthquake based on MCS intensity VIII would cause significant increase of stresses in the bridge. 1993.• • the value of the bending moment My is not negligible if compared to the moment due to static loads. These were used as input in the dynamic analysis of the bridge. J. ACNOWLEDGEMENTS Funding by NATO (Research Grant No.R. The modal shapes inferred from observed data are very similar to those obtained from modal analyses using the mathematical model. ASCE. One of the characteristics of the dynamic behaviour of the Indiano Bridge is the presence of longitudinal vibrations of high amplitude. & Huston D. & Bifulco A. 17(78): 485493. R. GardnerMorse M. one of the end bearings could be a fixed hinge. “Experimental modal analysis of the Garigliano cablestayed bridge”. MuriàVila D. Lecce L. Gomez R & King C. Earthquake Engineering and Structural Dynamics. 117(11): 33963416. while the bending moment Mz is about twice that one due to wind action. Clemente P. 4. the study of the seismic hazard in Florence site produced the uniform response spectra in terms of pseudovelocities. are very high. • Even thought Florence is not a high intensity seismic site. ASCE. Marulo S. “Dynamic Structural Properties of CableStayed Tampico Bridge”. both in vertical and in longitudinal directions. The authors wish also to acknowledge the Technical Office of the Florence City Hall for having allowed this study. Elsevier Science. carried out in terms of response spectrum analysis. stresses in the piers due to horizontal forces). 1995.. Int. Journal of Structural Engineering. 119(11): 33843404. REFERENCES 1. the axial forces in the tower and in the cables are much lower than the corresponding one due to static loadings. “Modal Identification of CableStayed Pedestrian Bridge”. Journal of Structural Engineering. 3.G. but in this case the above occurrence is amplified because of the absence of side cablestayed spans and because the system is not selfanchored. CRG 960109) is gratefully acknowledged. 1991. so that to determine frequent needs of maintenance works.
QR of RTRI. Geologie der Erdbeben. Clemente P. 16. 6(1): 145152. Ministero Lavori Pubblici “Aggiornamento delle norme tecniche per la progettazione. 11. & Paciello A. 14. MercalliCancaniSieberg macroseismic scale”. 4 maggio1990”. G. & Tunomoto M.U. “Engineering seismic risk analysis”. 404. Journal of Structural Engineering. Tab. 1993. 17. 1991. 2: 314 (in Italian). . 119(4): 10151031. Perilli N. Meletti C.. “Analisi sperimentale di un ponte strallato”. Pugliese A. Repubblica Italiana. Nakamura Y..T... 1930. Milano. Scandone P. C. 13. 9. & Huang D. D”Apuzzo M. Giannini Napoli. “Stima degli spettri di risposta da registrazioni di forti terremoti italiani”.PP. Am. Journal of Earthquake Engineering. 1992. Molin D. Patacca E. “Estimation of Damping Ratio of CableStayed Bridges for Seismic Design”.5. D’Apuzzo M.. 18. evoluzione cinematica e schema sismotettonico della penisola italiana”. 15. 77(5): 14911513. No. AutostradeItalstat Roma. “MCS. “Struttura geologica.. Wang T. Am. GNDT. 3(4): 475494.. Atti Convegno GNDT (Pisa. Bellatalla M. Proc. 1989. Kawashima K. 3: 7078 (in Italian). Clemente P.. “Attenuation of peak horizontal acceleration and velocity from Italian strongmotion records”. 10. Proc. Journal of Structural Engineering. & Sabetta F. ASCE. 1989. 30(1): 2533. Kawashima K. Seism. From Sieberg A. 29 Jan... 1987.. 118(5): 13541374. 1990. Augusti G. 1991. No.L.. 6. & Unjoh S.. 12. Seism Soc. Cornell C. LL. Unjoh S. Bull. & Santini U. 1999. & Pugliese A. 1968. “Progetto preliminare dei ponti strallati”. Imperial College Press. 162 (in Italian). Vignoli A. 7th Italian Conference on Steel Structures (Torino 1979). 58(5): 15831606. 2 and 3. 8. Berlin. Handbuch der Geophysic. ASCE. “A method for Dynamic Characteristics estimation of Subsurface using Microtremor on the Ground Surface”. Sabetta F. A. Japan Society of Structural Engineering and Earthquake Engineering. 7. “Damping Characteristics of CableStayed Bridges Associated with Energy Dissipation at Movable Bearings”. 1989. Soc. Bull. Ingegneria Sismica.M..A. Autostrade. 1. 1990). “CableStayed Bridge Vibration due to Road Surface Roughness”. 1990. Chiarugi A. “Analisi del modello generalizzato di ponte strallato” Fondazione Politecnica per il Mezzogiorno d”Italia. “Seismic Hazard Assessment in Florence City Italy”. 1979 (in Italian). l’esecuzione e il collaudo dei ponti stradali .D.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue reading from where you left off, or restart the preview.