Experimental Study of Seismic Performance on A Continuous Bridge with CFST Composite Truss Girders and Lattice High Piers

jumpclaybrainedUrban and Civil

Nov 25, 2013 (3 years and 9 months ago)

109 views

Experimental Study of Seismic
Performance

on
A Continuous Bridge with
CFST Composite Truss Girders and Lattice High Piers

Y.F. Huang
1
,
B.Briseghella
2
,
C.S.Liu
2
, Q.
X. Wu
2
, B.C. Chen
2

(1.
Department of Construction,
Università IUAV di Venezia
, Venic
e,
Italy
, 30123
)

(2. College of Civil Engineering, Fuzhou University, Fuzhou, Fujian, 350108)

Corresponding Author:

Yufan Huang; Email:
huangyufan325@gmail.com
; Address:
Department of Construction,
Università

IUAV di Venezia
,
Dorsoduro 2206
, Venice,
30123, Italy


Abstract

The Ganhaizi Bridge is located at a seismic area in Sichuan Province, China, only 480 km south
from Wenchuan where a deadly earthquake occurred on May 12, 2008, measured at 8.0 Ms and 7.9 Mw.

The
girder of Ganhaizi Bridge is a rigid frame continuous CFST truss girder, and its pier is a high CFST lattice
structure. Due to the light self
-
weight of those composite structures, the seismic performance is expected to
be improved significantly.

In or
der to study the seismic performance of the upper pier (H = 107m) of the bridge, a 1:8 scale model with
two spans was designed for multi
-
shaking tables test. The tests had been carried out by using artificial waves
based on Chinese code. The loading cases
of longitudinal input, transverse input and the combination of
transverse and longitudinal input had been considered.

In this paper, model design and some test results were presented.
D
ynamic characteristics and response of
the bridge model, such as fundam
ental frequency, acceleration, displacement

and
strain were measured

and analyzed.
Results showed that this new type of bridge ha
d

a favorable seismic performance.

Keywords
:

CFST truss; high pier; multi
-
shaking tables test; seismic performance.


1.
Introduction

In many recent earthquakes such as the 1994 Northridge (USA), 1995 Kobe (Japan) and 2008
Wenchuan (China) earthquakes, bridge collapses due to insufficient seismic performance of piers have
been found out
[1
-
3]
. At the same time, concrete fill
ed steel tubular (CFST) columns are applied more
frequently in recent years, especially in seismic regions. This is due to its excellent earthquake
-
resistant
properties, such as high capacity, favorable ductility and large energy absorption capacity
[4
-
6]
.

CFST
types may provide columns with higher bearing capacity and improved ductility since the concrete
prevents local buckling of the steel tube and the steel tube provides confinement to the concrete and
prohibits concrete spalling
[7].

The ductility of C
FST has been mostly studied for slender CFST
columns belonging to tall buildings located in seismic areas
[8
-
10].

In bridges, studies have been
performed for evaluating the ultimate strength and ductility of square and rectangular CFST piers
subjected to s
eismic lateral loads
[11]
. Moreover, the behavior of CFST has been widely studied
through hundreds of tests on CFST subjected to axial and flexural loads are nowadays available in the
literature

[12
-
14]
. These tests were carried out on CFST columns having
different slenderness ratios,
different steel sections and different concrete and steel strengths.
Extensive researches on modeling of
CFST trusses have already been made elsewhere, including investigation into local buckling in the steel
tube, and evaluat
ion of the strength and rigidity of truss connections
[15]
. This structure typology can be
also used to built
multi
-
storey frame building with excellent seismic resistance, and the system may be
practical
[16]
.

During

the recent construction activity

in
China, a

new
bridge
typology
with CFST composite truss
girders and lattice high piers has been
adopted
.
At present, however, some crucial aspects of
its
seismic
response are not fully understood, because of the great uncertainties associated with seismic a
ction
modelling. In fact, typical capacity design strategies are more likely to fail in those bridges whose
seismic
behavior
is governed by tall pier dynamics

[17]
.

I
t is a main way to use shaking table test to study the seismic performance of the structu
re.
In the
present study,

the se
ismic performance of this new system is
investigated
.

The research results on a 1:8
scale model with two spans and three
columns
, designed and tested on
the
multi
-
shaking tables at the
State Key Laboratory in Civil Engineeri
ng, Fuzhou University, P. R. China.
A
rtificial waves based on
Chinese
C
ode
regulations
were
used

for the seismic test
s

and

fundamental frequency,

acceleration,
displacement and strain of the model were measure
d and analyzed
.

2.
C
ase
s
tudy

The
subject of

this research is
the
Ganhaizi Bridge,
which was built in 2011 and is
located in Sichuan
Province, China, 480 km south from Wenchuan where a deadly earthquake occurred in 2008, measured
at 8.0 Ms and 7.9 Mw

(
Fig.1
.
).



Fig
.
1
.

Ganhaizi Bridge

The Ganhaizi

Bridge has a total length of 1811 m
. It is

composed
of

three continuous truss girder units
with a span arrangement of 40.7 m + 9×44.5 m + 40.7 m, 45.1 m + 4×44.5 m + 45.1 m and 45.1 m +
3×44.5 m + 11×62.5 m + 3×44.5 m + 45.1 m
.
Fig
.
2

illustrates the long
est one.


Fig
. 2.

The elevation of the longest unit of the Ganhaizi Bridge (unit: cm)

Two

types of piers

were used
: (a) reinforced concrete (RC) piers for
height less than 25 m

and (b) CFST
piers for
height more than 25 m

(the tal
lest one is 107 m). The CFST pier is generally composed of four
CFST

columns

connected together by steel truss tubes. The CFST columns have a diameter of 813 mm
or 720 mm with the wall thicknesses from 12 mm to 16 mm, filled with C50 concrete. For the pier
s
taller than
90

m

(No.16~25 piers)
, the truss tubes are replaced with 40 mm thick RC
webs

at the bottom
region of 30 m height to enhance its stiffness
. Slant supports are added on top of CFST columns, fixed
with girder together, shown in
Fig.
3
.

The super
structure comprises of concrete bridge deck, steel truss web and CFST bottom chord. The
cross
-
section of the truss girder with 440 cm height is shown in
Fig.
4
.

The bottom chord tubes are
CFST with a diameter of 813 mm and thickness from 18 mm to 32 mm, filled with C60 concrete. The
web steel tubes have a diameter of 406 mm.

Research prototype

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
3×4450=13350
4510
11×6250=68750
3×4450=13350
4510
The use of CFST in piers facilitates the construction and improves the ductility of the s
tructure. The
composite structure
allows to take the most of
steel and concrete

material properties. Therefore, the light
CFST structures can reduce the seismic response, save materials and facilitate construction.


Fig. 3. CFST composite piers (unit: cm)

CFST bottom chord
tubes (
φ
813)
Web steel tubes
(
φ
406)
1225
1075
200
1075
1225
1225
440
28
70
50
50
Steel fiber reinforced
concrete deck

Fig
.

4.

Cross
-
section of truss girder (unit: cm)

3.
Design and Manufacture of
Test

Model

3.
1

M
odel
size
design

Considering the model manufactured, it
was

not easy to fill concrete into steel tubes for large scale
proportion model. Meanwhile, subjected to the restrictions of the length and bearing capacity of the test
device, the scale ra
t
io of the girder and piers
was

determined
as
1:8. The elevation drawi
ng of the bridge
model is shown in
Fig.
5
. The footing was rigidly connected to the multi
-
shaking tables through
high
-
strength bolts.

3.
2

Dynamic similitude

relationship design

Dynamic similitude theory was adopted to make sure that the model ha
d

a similar
behavior

to the
real bridge
.
According to

π

terms
[18]
, the similarity ra
t
io of corresponding physical quantities between model and
prototype satisfies the condition as follows
,


153
18.75
38
38
73
38
38
173
306
73
1363
15
251
416
416
251
15
55
101
125
25
10.1
31×25
106
25
125
25
383.65
1337.5

Fig. 5.

Elevation drawing of the bridge model

(unit: cm)

Table 1 Similitude relation of physical quantities

Physical quantity

Scaling law

Scaling factor

Linear Length


1/8

Displacement


1/64

Stain


1/8

Elastic modulus


1

Density


1

Frequency


1/8

Time

,

1/8

Acceleration


1

Table 2
Material properties

Steel tube

Elastic
modulus
(MPa)

Yield
strength
(MPa)

Ultimate
strength
(MPa)

Concrete

Elastic
modulus

(MPa)

Axial
compressive
strength

(MPa)

Steel web

2.00×10
5

314

535

Bottom chord

2.71×10
4

37.1

Bottom chord

2.00×10
5

364

501

Deck

3.96×10
4

52.7

Column

2.02×10
5

375

465

Column

3.23×10
4

42.3

Truss
tubes

2.03×10
5

388

502

RC web

3.15×10
4

24.4

Slant support

2.00×10
5

380

497




In this research,
a
s far as possible the same material
s used in the real case were chosen and
no mass on
the model

were added
. Input acceleration
intensity was

the same as prototype
, only the length was
adjusted
.
For tall pier,
seismic
performance evaluation

mainly concerned about
displacement on the top,
which is

acceleration
quadratic integral
to
wards time.
Table 1

list the s
imilitude relation of physical
quantities
.
The height of the model
was

13.9

m and total mass
20.9 ton.
Measured material properties
are
listed in
Table 2
.

Longitudinal
accelerometer
Transverse
accelerometer
Displacement
transducer
2(4,6)
1(3,5)
Unidirectional
strain gauge
Bidirectional
strain gauge
1(2)
3(4)
5(6)
7
8
1-1
1-2
1-3
1-4
1-5
2-1
2-2
2-3
2-4
2-5
3-1
3-2
3-3
3-4
3-5
1-6
1-7
1-8
1-9
1-10
2-6
2-7
2-8
2-9
2-10
3-6
3-7
3-8
3-9
3-10
Shaking
table
Shaking
table
Shaking
table
1-2,1-3
1-4,1-5
1-6,1-7
1-10(1-20)
1-1(1-11)
(1-12,1-13)
2-2,2-3
2-1(2-11)
(2-12,2-13)
3-2,3-3
3-1(3-11)
(3-12,3-13)
(1-14,1-15)
2-4,2-5
(2-14,2-15)
3-4,3-5
(3-14,3-15)
(1-16,1-17)
1-9(1-19)
1-8(1-18)
2-6,2-7
2-10(2-20)
(2-16,2-17)
2-9(2-19)
2-8(2-18)
3-6,3-7
3-10(3-20)
(3-16,3-17)
3-9(3-19)
3-8(3-18)
1
11
2,3
12,13
14,15
4,5
18,19,20
16,17
8,9,10
6,7

Fig
. 6.

I
nstrumentation

arrangement

details


(a)
A
ccelerometers


Fig. 8.

Panorama



(b)
D
isplacement transducers

Fig. 7.

Transducers layout

3.
3 Instruments and transducers

The motion of the shaking table and of the model were monitored through a variety of instrumentation
s

composed by

30 accelerometers, 8 displacement transducers and 60 strain gages, including longitudinal,
transverse and vertical directions. A
high
-
speed camera was used to measure the displacements.
Displacement transducers were installed at the top of the deck and co
lumns. Since the stress
es

w
ere

expected to concentrate at the piers where the stiffness varies, strain gages were mainly
placed

at the
bottom of columns and RC web, and piers connect with slant support and RC web. Accelerometers were
placed at the top of d
eck, columns connect
ed

with slant support and RC web, cent
er

along height and
top of shaking tables.

The

data
of each test
were recorded simultaneously through instruments and
transducers.

Instrumentation arrangement details

are
shown in
Fig.

6
and

Fig.

7.

The final model for
shaking table test can be seen in
Fig.
8
.

4
.
Dynamic characteristics

analysis

4.1
Dynamic
characteristics

of experimental model

Using

the white noise excitation and
fast Fourier transform (FFT)
, the fundamental frequency of the
structure
was

identified.
Fig. 9

and
Table 3

show

the FFT
of
deck accelerations and displacements for
white noise excitation in longitudinal and transverse direction
considered
individually and
the
combination of two
directions. It was found that the first longitudinal frequency and transverse
frequency
were

1.45Hz and 2.10Hz, respectively.
Damping of each mode was calculated through
h
alf
-
power bandwidth method
.

0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.00
0.01
0.02
0.03
0.04
0.05
1.45 Hz
PSD (m
2
/s
3
)
Frequency (Hz)
FFT of Deck Acc.
Longitudinal
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0
0.4
0.8
1.2
1.6
2.0
1.45 Hz
PSD (m
2
/s
3
)
Frequency (Hz)
FFT of Deck Disp.
Longitudinal

(a)
FFT, longitudinal direction

0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.00
0.01
0.02
0.03
0.04
0.05
2.10 Hz
FFT of Deck Acc.
Transverse
PSD (m
2
/s
3
)
Frequency (Hz)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.00
0.04
0.08
0.12
0.16
0.20
2.10 Hz
FFT of Deck Disp.
Transverse
PSD (m
2
/s
3
)
Frequency (Hz)

(a)
FFT, transverse direction

0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.00
0.01
0.02
0.03
0.04
0.05
1.45 Hz
PSD (m
2
/s
3
)
Frequency (Hz)
FFT of Deck Acc.
Longitudinal
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.00
0.01
0.02
0.03
0.04
0.05
2.10 Hz
FFT of Deck Acc.
Transverse
PSD (m
2
/s
3
)
Frequency (Hz)

(c
)
FFT,
c
ombination of two directions

Fig. 9. FFT analyses of white noise excitation

Table
3 White noise excitation results

Order

Frequency


Hz


Damping

Modal shape

1

1.45

0.014


2

2.10

0.019




4.2
Dynamic
characteristics

of
Ganhaizi Bridge

A three
-
dimensional finite element model of the
Ganhaizi Bridge

was implemented in Midas/Civil, shown in
Fig
.
10
. The concrete web plates at the bottoms of the composite columns were simulated by plate elements,
while beam elements were used for all the others parts of the bridge, for a total number of 14443 nodes and

24441 elements. The columns bases were considered as fix in all degre
e
s of freedom. According to the
design documents, columns from No.15 to No.26 were fixed with girder, others were connected using spring
element, which the value of stiffness list in
Tab
le
4
.
The subspace method
and Newmark time integration
method (
β = 0.25, γ=0.5
) were

adopted

for eigenvalue and seismic analysis, respectively. The damping
coefficients of all members were
uniformly

set to 0.02 and Rayleigh damping was used. T
he first
mode
s

in
plane and out of plane
with
large effective masses
were identified and
used for Rayleigh damping. The initial
stress
es

on the bridge

were

assumed to be the stress
es

obtained under the dead load condition
[19]
.


Fig. 10.

F.E. model of
Ganhaizi Bridge

Table
4

Stiffness of rubber bearing (unit: kN/m)

Number of column

11

12

13

14

27

28

29

30

Vertical direction

Fix

Fix

Fix

Fix

Fix

Fix

Fix

Fix

Longitudinal direction

1000

2400

2400

2400

2400

2400

2400

1000

Transverse direction

Fix

Fix

Fix

Fix

Fix

Fix

Fix

Fix


Table 5

shows the frequency comparison between the real bridge and the test model. Results showed that the
radio was 1:7.47 in first natural frequency and 1:7.69 in second natural frequency, which were close to the
theoretical frequency ratio of 1:8

and it demonstrated the accuracy

of similitude relationship.


Table

5

Frequency comparison

Mode

Prototype

(Hz)

Modal Shape

Test Model

(Hz)

Test Model /

Prototype

1

0.
194


1.45

1:
7.
47

2

0.2
73


2.
10

1:7.69


5. Experimental procedure and results

5.1
Seismic wave and verification

The tests had been carried out by using artificial waves based on Chinese code and
generated according to
the design response spectrum of
the real bridge.

T
he
design

seismic
acceleration
action

was 0.362g
associated with a reference probability of exceedance, P
NCR
=10%

in 50 years
.
Durable time of artificial
wave was

adjusted
through

similitude relation.
Fig.
11

show
ed

the design response spectrum and
artificial wave for the model and prototype, fundamental
period of the model and prototype are
indicated with a dashed line.
Fig. 12

and
Fig. 13

show
ed

the artificial waves for prototype and model,
which
fitted from design spectrum. T
he peak ground acceleration (PGA)
was

0.16g.

Fig.14
list the
displacement compa
rison on the top of column between the model and the prototype under PGA=0.16g
with different durable time, respectively. The ratio between the model and prototype was close to 1:8
2
. It
showed that the test results correctly reflected the real displacement

of prototype.

0.01
0.1
0.5
1
2
5
10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Model T=0.69s
Artificial wave
(PGA=0.16g)
prototype
model
Design spectrum
prototype
model
Acceleration (g)
Natural period (s)
Prototype T=5.15s

Fig
. 11.

Response spectrum of input
motions


0
5
10
15
20
25
30
-2
-1
0
1
2
Acc. (m/s
2
)


Time (s)
PGA=0.16g

Fig. 1
2.

Artificial wave for
prototype

0.000
0.625
1.250
1.875
2.500
3.125
3.750
-2
-1
0
1
2
Acc. (m/s
2
)


Time (s)
PGA=0.16g
Fig. 13
.

Artificial wave for experiment

0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
3.25
3.50
3.75
-0.006
-0.004
-0.002
0.000
0.002
0.004
0.006
Test Disp.(m)


Time(S)
Prototype Disp.(m)
Prototype Disp.
Test Disp.


0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
-0.4
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
3.25
3.50
3.75
-0.006
-0.004
-0.002
0.000
0.002
0.004
0.006


Time(s)
Prototype Disp.(m)
Prototype Disp.
Test Disp.(m)


Test Disp.

(a) Longitudinal displacement (b) Transverse displacement

Fig. 14. Displacement comparison

5.2 Test procedure

During the test, PGA
was adjusted from 0.05g to the maximum intensity that the shaking table could
aff
ord. The
m
axim
um

overturning moment of
the
big table

was

50
~80t∙m, while that of the small one
was
20
~
30t∙m
, so the
artificial wave
finally
include
d

0.05g, 0.10g, 0.15g, 0.16g, 0.20g, 0.22g, 0.24g,
0.26g, 0.28g, 0.30g, 0.35g, 0.40g, 0.45g, 0.50g

in longitu
dinal direction, while
maximum wave in
transverse direction
wa
s PGA=0.22g. The loading cases of longitudinal input, transverse input and
combination input of both
were

also considered.

5.3

Response characteristics for transverse excitation

The assessment of the response
wa
s based on the envelope of the maximum displacement and acceleration
values
. When the model was subjected to transverse excitation,
Fig. 15

showed the maximum longitudinal
and transverse displacement on the top of center pi
er. Transverse displacement gradually increased as the
increasing acceleration intensity, but longitudinal displacement was small.
Fig. 1
6
showed the maximum
transverse acceleration of center column. It can be seen that the value of deck was nearly the sam
e as that on
shaking table. In the lattice column zone, acceleration significantly magnified, and reduced the acceleration
on the deck through remarkable
oscillation

on the columns.

Table 6

list the maximum strain in transverse excitation. The column was u
pright along transverse,
therefore, strain increased from top of column to the footing. Due to the RC web shared the internal
force, position of maximum strain was on the top of RC web.

0.0
0.1
0.2
0.3
0.4
0.5
0
3
6
9
12
15
Longitudinal Disp.
Transverse Disp.
Disp. on the top of column (mm)
PGA (g)

0.0
0.1
0.2
0.3
0.4
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Deck (2-6)
Slant support (2-7)
Centre (2-8)
Top of RC web (2-9)
Shaking table (2-10)

Acc. (g)
PGA (g)

Fig. 15. Transverse displacement Fig. 16. Transverse acceleration

Table 6 Maximum strain in transverse excitation (PGA=0.22g, unit:
)

Position

Vertical strain

Slant support

37.83

Slant support of column

45.16

Top of RC web of
column

54.32

Bottom of column

45.78

Concrete on bottom of RC web

29.91

Steel bar on bottom of RC web

48.22


5.
4 Response characteristics for longitudinal excitation

When the model was subjected to longitudinal excitation, experimental response was the same as mentioned
above.
Longitudinal displacement increased, and transverse displacement was
not
affected significant
,
shown

in
Fig. 17
.

M
aximum longitudinal accelerati
on of center column
wa
s presented in
Fig. 1
8
, which shows
a
favorable seismic performance

of CFST composite columns.

Table
7

lists the maximum strains in longitudinal excitation. The maximum strains on columns were in
the position of slant support and top
of RC web, where stiffness of columns changed. The vertical
strains were significantly larger than longitudinal strains, but far less than yield strain.

0.0
0.1
0.2
0.3
0.4
0.5
0
3
6
9
12
15
Longitudinal Disp.
Transverse Disp.
Disp. on the top of column (mm)
PGA (g)

0.0
0.1
0.2
0.3
0.4
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Deck (2-1)
Slant support (2-2)
Centre (2-3)
Top of RC web (2-4)
Shaking table (2-5)

Acc. (g)
PGA (g)

Fig. 17. Longitudinal displacement


Fig.18. Longitudinal acceleration

Table 7 Maximum strain in longitudinal excitation (PGA=0.22g, unit:
)

Position

Longitudinal strain

Vertical strain

Slant support



31.74

Slant support of column

32.96

117.19

Top of RC web of column

17.70

98.88

Bottom of column

13.12

84.88

Concrete on bottom of RC web



47.61

Steel bar on bottom of RC web



31.07


5.
5 Response characteristics for
bidirectional

excitation

W
hen the model
was

simultaneously subjected to bidirectional seismic inputs, the displacement
s

and
strains were
not larger than
that obtained from the
uni
dire
ctional seismic input
s
, see
Table
8
and

Table 9
.
Therefore, it was not necessary to consider the influence of
bidirectional inputs. E
xcitation from
transverse direction ha
d

greatly influence than longitudinal excitation.

According to Ref.

[20]
, the
horizontal displacement on top of piers should not exceed 1/300 e
levation

of piers under design
response spectrum. Th
e elevation of prototype was 107m (No.20 pier), therefore, the displacement
limitation was 357mm.
Table
10

illustrates the displacement verification, showing that the displacement
was within limit values.

Table 8 Summary of maximum displacement (PGA=0.22g
, unit: mm)

PGA

Longitudinal Disp. (mm)

Transverse Disp. (mm)

Unidirectional

(1)

Bidirectional

(2)

(2)/(1)

Unidirectional

(1)

Bidirectional

(2)

(2)/(1)

0.05

0.874

0.931

1.065

1.166

1.161

0.996

0.10

1.617

1.959

1.212

2.137

2.239

1.048

0.15

2.962

2.877

0.971

3.518

3.626

1.031

0.16

2.742

3.317

1.210

4.062

3.760

0.926

0.18

3.338

3.297

0.988

4.183

4.315

1.032

0.20

3.448

5.963

1.729

5.305

5.577

1.051

0.22

5.594

6.963

1.245

5.090

5.457

1.072


Table 9 Vertical maximum strain in different directional
excitation (PGA=0.22g, unit:
)

Position

Longitudinal
excitation

Transverse
excitation

Bi
directional
excitation

Slant support of column

23.80

34.15

40.28

Top of RC web of column

26.25

43.95

43.33

Bottom of column

23.20

54.93

54.32


Table 10
Displacement verification (PGA=0.16g, unit: mm)


Test results

Converted to
prototype

Limit values

Longitudinal Disp.

2.742

175

357

Transverse Disp.

4.062

260

357

6
.
Conclusions

Due to the ductility of CFST materials and
light
-
weight structural form
,
this new type of bridge exhibits
a favorable seismic performance. B
ased on a continuous bridge with CFST composite truss girders and
lattice high pier as prototype,
a 1:8 scale two
-
span model was designed for multi
-
shaking table test.

Seismic behavior of t
he bridge model was investigated under
artificial waves.
This paper introduces the
experimental results and the following conclusions can be drawn:

(1)

Through white noise excitation,

the identified fundamental frequency of the structure was 1.45 Hz in
lo
ngitudinal direction and 2.10 Hz in transverse direction.
The ratio of frequency between prototype and
model was 1:7.47 in first order and 1:7.69 in second order, which were closed to the theoretical
frequency ratio of 1:8 and verified the accuracy of simi
litude relationship.
The
model
displacement
s

also showed a good agree
ment
to the prototype according to the similitude relationship.

(2)

Under transverse excitation,
acceleration significantly magnified and reduced response on the deck
through remarkable
oscillation on the columns. The maximum strain on the column was on
the

top of
RC web.

(3)

Under longitudinal excitation,
experimental
response
was nearly the same as under transverse
excitation.

The maximum strain
s

on columns w
ere

in the position of slant

support and top of RC web,
where stiffness of columns changed. The vertical strain
s

w
ere

significantly larger than longitudinal
strain, but less than yield strain

so

the model
remained in the elatic range
.

(4) Under combination of longitudinal and transve
rse direction, the displacement
s

and strain
s

were

not

larger than subjected to one directional seismic input. Therefore, it was not necessary to consider the
influence of bidirectional inputs. Excitation from transverse direction had greatly influence than

longitudinal excitation.

(5) Displacement verification was within limit values
as proposed by

[20]
. Finally t
his new type of
bridge
seems to
ha
ve

a favorable seismic performance
.

R
eferences

[1] Broderick BM, Elnashai

AS. Analysis of the failure of interstate 10 freeway ramp during the
Northridge earthquake of 17 January 1994. Earthq Eng Struct Dyn 1995;24:189

208.

[2] Elnashai AS, Bommer JJ, Baron CI, Lee DH, Salama AI. Selected engineering seismology and
structural e
ngineering studies of the Great Hanshin earthquake of 17 January 1995. ESEE research
report: no. 95
-
2. London: Imperial College; 1995.

[3] Lee G. The 512 Wenchuan Earthquake of China

a preliminary report. Department of Civil,
Structural and Environmental E
ngineering, MCEER; June 16, 2008.

[4] Lin
-
Hai Han, Wei Li. Seismic performance of CFST column to steel beam joint with RC slab:
Experiments. Journal of Constructional Steel Research 66 (2010):1374
-
1386.

[5] ASCCS. Concrete filled steel tubes

a comparison of international codes and practices. ASCCS
seminar, Innsbruck, Austria; 1997.

[6] Shanmugam NE, Lakshmi B. State of the art report on steel

concrete composite columns. J Constr
Steel Res 2001;57(10):1041
-
80.

[7] Huiling Zhao, Sashi K. Kunnath, Yong Yuan. Simplified nonlinear response simulation of composite
steel
-
concrete beams and CFST columns, Engineering Structures 32 (2010): 28252831

[8] Varma A, Ricles J, Sause R, Lu L. Seismic behavior and modeling
of high strength composite
concrete
-
filled steel tube (CFT) beam
-
columns. Journals Construct Steel Res 2002; 58(5

8):725

58.

[9] Zubydan A, ElSabbagh A. Monotonic and cyclic behavior of concrete
-
filled steel
-
tube
beam
-
columns considering local buckling eff
ect. Thin
-
Walled Structure 2011; 49(4):465

81.

[10] Mao X, Xiao Y. Seismic behavior of confined square CFT columns. Engineering Structure
2006;28(10):1378

86.

[11] Kitada T. Ultimate strength and ductility of state
-
of
-
the
-
art concrete
-
filled steel bridge p
iers in
Japan. Eng Struct 1998;20 (4

6):347
-
54.

[12] El
-
Tawil S, Deierlein GG. Strength and ductility of concrete encased composite columns. Journal of
Structural Engineering, ASCE 1999;125(9):1009

19.

[13] Dundar C, Tokgoz S, Tanrikulu AK, Baran T. Behavi
or of reinforced and concrete
-
encased
composite columns subjected to biaxial bending and axial load. Building and Environment
2008;43:1109

20.

[14] Shanmugam NE, Lakshmi B. State of the art report on steel
-
concrete composite columns. Journal
of Constructio
nal Steel Research 57 (2001): 1041

80.

[15] Galambos TV. Guide to stability design criteria for metal structures. 5th ed., New York: John Wiley
& Sons, Inc, 1998.

[16] A. Kawano, K. Sakino. Seismic resistance of CFT trusses, Engineering Structures 25 (2003
):
607

619.

[17] Rosario Ceravolo, Giacomo Vincenzo Demarie, Luca Giordano, Giuseppe Mancini, Donato Sabia.
Problems in applying code
-
specified capacity design procedures to seismic design of tall piers.
Engineering Structures 31 (2009):1811
-
1821.

[18] Har
ry G. Harris, Gajanan M. Sabnis, Structural modeling and experimental techniques (second
edition), CRC Press, 1999:70
-
77.

[
19
]
Ren W.X and Obata M, Elastic
-
plastic seismic behavior of long span cable
-
stayed bridges, Journal
of Bridge Engineering, ASCE 1999
, 4(3): 194
-
203.

[
20
] Sichuan Provincial Communications Department Highway Planning Survey and Design Institute, (2008),
Guide to design and construction technology of road steel tube concrete bridge (in Chinese), China
Communications Press, Beijing, China
.