ENHANCED MODELS FOR THREE-DIMENSIONAL ANALYSIS OF REINFORCED

CONCRETE TWO-BLADE SLENDER BRIDGE PIERS

Sgambi Luca
1)
and Bontempi Franco
2)


Department of Structural and Geotechnical Engineering,
University of Rome La Sapienza, Italy




ABSTRACT

A modern conceptual design of a bridge structure should be open to wider criteria, which assure
that the structure is endowed with static, dynamic and ductile characteristics sufficient to tackle
the seismic events. To the satisfaction of these requirements and to the exploration and to the
verification of innovative structural schemes, one must consider refined and subtle analysis
formulations and tools.
Such a kind of structural scheme is based on the two-blade slender bridge piers. These
are elements subdivided into two parts with different geometric and mechanical properties. The
first part has a box section, highly stiff; two flexible blades connected at the top compose the
second part. The goal of this work is to show the necessity to develop a three-dimensional
analysis, the only one able to put in evidence some peculiar structural behavior.

INTRODUCTION

In the Conceptual Design of bridges, in areas of moderate to high seismicity, the selection of
ductile behavior is a general strategy inside the so-called Integral Bridge Design (Biondini 2001,
Calzona and Bontempi 2001). Its implementation, either by providing the formation of a
dependable plastic mechanism or by using base isolation and energy dissipating devices, must be
decided. When a ductile behavior is selected, the following main points should be considered.
a) The number of supporting elements (piers and abutments) that will be used to resist the seismic
forces in the longitudinal and in the transverse direction must be decided. In general
continuous structures behave better under earthquake conditions than bridges having many
movement joints. The optimum post-elastic seismic behavior is achieved if plastic hinges
develop approximately simultaneously in as many piers as possible. However, the number of
earthquake resisting piers may have to be reduced, by using flexible mountings between deck
and piers in one or in both directions, to avoid either high reactions due to restrained

1)
Ph.D. Student
2)
Professor
deformations or an undesirable distribution of the seismic actions and/or of the capacity design
effects.
b) A balance should be maintained between the strength and the flexibility requirements of the
horizontal supports. High flexibility reduces the level of the design seismic action but
increases the movement at the joints and moveable bearings and may lead to high second
order effects.
The two-blade pier is a structural scheme that can furnish the right structural restoring (Fig.1).
They have a double scope. First, it is the transmission of vertical loads to the foundation
structures, second, it is the possibility to allow horizontal displacements (thermal effects,
concrete shrinkage, etc.) without the use of cinematic mechanisms. The ratio between the low
rigid box section and the high flexible two-blade section characterizes the structural behavior
with respect to the horizontal loading system. If the rigid box section has a small height respect
the total pier height, the behavior will have great ductility, but little resistance. Instead, if the
rigid box section is dominant, it increases the stiffness with respect to the horizontal loads but it
decreases the ductility. The choice of the relative height of the rigid box section is of great
importance because this has a great influence on the global structural behavior.
A non-linear analysis is needed to show the global behavior of these structures. In this work a
material non-linear analysis is developed. Shear behaviors and diffusion effects will also be
considered, using a three-dimensional model. An interesting comparison, between the non-linear
three-dimensional model and non-linear beam model will be showed.

Longitudinal section


Fig. 1. Longitudinal section of the bridge with (a) two-blade piers (unit = cm).


THE PROBLEM DEFINITION

The structure that has been analyzed is reported in Fig.1 (Biondini 2001). The pier is
composed of two parts, the first part (of height h
1
) with a rigid box section, and a second part (of
height h
2
) composed of two flexible blades. The a-dimensional ratio = h
1
/ (h
1
+h
2
) is used to
characterize different piers. The pier has been modeled using eight nodes finite elements (brick)
and the structural response has been analyzed for the ratio = h
1
/ (h
1
+h
2
) = 0/3, 1/3, 2/3, 3/3.
Two-blades pier
h
1

h
2

4
0
0
0

5000 5000
A three-dimensional generalization of membrane constitutive law (CFT, MCFT, RA-STM
etc.) has been implemented to study the material non-linear behavior (Vecchio and Collins
1988). The constitutive law is a rotating fissure and the non-objective mesh is also considered
using a non-local damage regularization (Bontempi and Malerba 1997). The uniaxial law for
the compression concrete is:
1
1
]
1

,
_

,
_

2
0 0
2 '

d d
c d
f if ( )
0

d
(1)
1
1
1
1
1
]
1

,
_

2
0
1
2
1
1 '

d
c d
f if ( )
0
>
d
(2)
where, the softened parameter
eq

400 1
1
is used to reproduce the compression resistance if
there is a transversal tensile strain. The equivalence strain is assumed as
2
3
2
2
2
1
+ +
eq

where
i
is zero if the
i
is of compression strain. The constitutive law for the tensile concrete
is:
r c r
E if (
cr r
) (3)
4 . 0

,
_

r
cr
cr r
f

if (
cr r
> ) (4)
where
3
9500
c c
f E is the initial elastic module, and the
3 2
25 . 0
c cr
f f is the cracking tension.

Fig. 2. Constitutive concrete laws.

The shear rigidity is assumed as:
j i
j i
ij
E E
E E
G
+

(5)
The rigidity matrix D is valued in the principal directions system for means of the uniaxial
laws Eq. (1-4). The rigidity matrix D
xyz
in the global reference system is:

T D T D
T
xyz

(6)
where T is a rotation matrix.
The reinforcement steel is assumed elastic perfectly plastic and his constitutive laws are:

s s s
E if
y s
< (7)
y s
f if
y s
(8)
The pier analyzed is 42.50 m height and the vertical load is V=30 MN for each blade. The
mechanical characteristics are: f
c
= -33.2 MPa, f
sy
= 430 MPa, E
s
= 210 GPa,
su
= 1%.
The increasing horizontal load is applied to the top of the pier. For the sake of notice, a
Pentium III

533 MHz has been used for the analysis, the calculation time has been 7 hours for
the 6000 degrees of freedom model.



Fig. 3. The pier discretization and the flexion planes.


RESULTS

Two directions of horizontal load have been considered. The first is parallel to the two blade
planes (direction force B in figure 3), the second (more interesting) is perpendicular to the two
blade planes (direction force A in figure 3).
In the figures 4, 6 and 7 one is resuming the analysis of case A, while in the figures 5, 8 and 9
the case B. In figure 5 the results of the case A are compared with a beam model analysis. In the
figures 6, 7 and 9 the crack orientations are plotted and the damage zones are qualitatively
represented.
In the case A, the ultimate load increases itself of 25%, by the increase of parameter , while
the structural ductility does not vary. The structural behavior is similar to the beam behavior,
with the crack zones near to the maximum momentum zones and the crack orientation
perpendicular to the axis of the beam. The discontinuity zone between the rigid box section and
the two-blade section causes a diffusion region. This diffusion zone, does not influence the
global structural behavior, and the response for the different parameters is similar.
Great variations are present in the case B, when the parameter changes. The ultimate
horizontal load for =1 is four times that for =0, and the ductility decreases. These variation
are caused by the great difference of behavior between the rigid box section and the two blades

Axis of the bridge
A
B
section when the flexion plane is the one of case A. In Fig.8 is showed the great deformation of
two-blades section with respect to the rigid box section: the horizontal displacements are
concentrated in the top of the pier. In the Fig.5 the results are compared with the beam model
analysis (Biondini, Bontempi, Malerba, and Martinez y Cabrera 2001). We note that for =0
and =1 the results are identical, while for the =1/3 and =2/3 the ultimate load calculated by
the beam model is 20% greater of the one calculated by three-dimensional model. The cracked
zones are showed in Fig. 9.














Fig. 4. Horizontal load-displacement curves for the case A.



Fig. 5. Horizontal load-displacement curves for the case B.

0
2
4
6
8
10
12
14
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Spostamento in sommit (m)

Parametro
crescente
Displacement at the top (m)
H
o
r
i
z
o
n
t
a
l

f
o
r
c
e

(
M
N
)

increased
0
2
4
6
8
10
12
14
16
18
0 0.
2
0.
4
0.6 0.8 1 1.2
Spostamento in sommit (m)
Forza orizzontale (MN)
Beam element
Brick element
increased
H
o
r
i
z
o
n
t
a
l

f
o
r
c
e

(
M
N
)

Displacement at the top (m)


Fig. 6. Crack and damage for the case A when the horizontal load is 50%, 75% and 100% of the
ultimate load.






Fig. 7. Crack and damage for the case A when the horizontal load is 50%, 75% and 100% of the
ultimate load.

Fig. 8. Discretization and resultant deformation for the case B with = 2/3.




Fig. 9. Crack and damage for the case B when the horizontal load is 50%, 75% and 100% of the
ultimate load.
CONCLUSIONS
A modern conceptual design of a bridge structure should be open to wider criteria that assure
that the structure is endowed with static, dynamic and ductile characteristics sufficient to tackle
the seismic events. It is worth noting that a correct conception of the structural morphology,
besides the achievement of better guarantees with regard to the seismic behavior, may also lead
to economic advantages either in the realization of the whole structural system, or in the
adoption of seismic devices, with a consequent possible reduction of the management cost of the
structure as well. In the context of a quality design, aimed to the achievement not only of
strength requirements but also of durability and functionality of the structure, one can also
consider the possibility to convoy the damage towards some regions of known location, possibly
planning also the modalities of the restore in the project itself.
The critical aspects that give clarity and coherence to the whole structural design need
sophisticated analysis formulation and tools. A non-linear analysis is needed to show the global
behavior of these structures. In this work, shear behaviors and diffusion effects are considered,
using a material nonlinear three-dimensional model.

REFERENCES
Biondini, F., (2000), Strutture da ponte soggette ad azioni di tipo sismico. Modellazione ed
ottimizzazione. Ph.D. thesis, Politecnico di Milano, (in Italian).
Biondini, F., Bontempi, F., Malerba, P.G. and Martinez y Cabrera F., (1999), Analisi non lineare
di pile da ponte a doppia lama, Giornate AICAP, Torino 4-6 novembre vol. 2 Pag. 103
112, (in Italian).
Biondini, F., (2001), Conceptual seismic design of damage-tolerant bridges., 7
th
International
Seminar on Seismic Isolation, Passive Energy Dissipation and Active Control of Vibrations
of Structures, Assisi, Italy, October 2-5, 2001.
Bontempi, F. and Malerba, P. G., (1997), The role of softening in the numerical analysis of R.C.
framed structures, Structural Eng. And Mech., vol. 5, 1997.
Bontempi, F., Malerba, P. G. and Romano L., (1995), Formulazione diretta secante dellanalisi
non lineare di telai in C.A./C.A.P., Studi e Ricerche, F.lli Pesenti, Politecnico di Milano,
16, 351-386, (in Italian).
Calzona, R., Bontempi, F., (2001), Remarks on the approval process of designs of structures
provided with innovative anti-seismic systems in Italy, 7
th
International Seminar on Seismic
Isolation, Passive Energy Dissipation and Active Control of Vibrations of Structures,
Assisi, Italy, October 2-5, 2001.
Gomez, C., (2001), Sulle differenze di capacit portante deducibili dalle teorie a fessurazione
diffusa. Master thesis, F.lli Pesenti, Politecnico di Milano, (in Italian).
Mancini, G., Martinez y Cabrera, F., Pisani, M. A. and Recupero A., (1998), Behavior of
Nonlinearly Restrained Slender Bridge Piers, ASCE J. of Bridge Eng., 3(3), 126-131.
Sgambi, L., (2001), Modellazione tridimensionale di strutture in C.A. e C.A.P. in campo non
lineare, Master thesis, F.lli Pesenti, Politecnico di Milano, (in Italian).
Vecchio, F. J., (1989), Non linear finite element analysis of reinforced concrete membranes, ACI
Structural Journal, January-February 1989, 26-35.
Vecchio, F. J. and Collins, M. P. C., (1988), Predicting the response of reinforced concrete
beams subjected to shear using modified compression field theory, ACI Structural Journal,
May-June 1988, 258-268.