Models for Simulating the Seismic Response of Concrete Structures - Dynamic Behavior of Concrete and Seismic Engineering - page 268

A 3D model based on multi-layer thin shell modeling has been adopted to
simulate the behavior of this structure. It should be noted that with such modeling,
transverse shear is not taken into account, and the actual 3D behavior (transverse
dilation) is not correctly described. Nevertheless, the approach allows the bi-bending
of the different parts of the wall and the shear in the plane of the flanges and girder
web to be represented quite well. The wall is modeled using integrated DKT-type
elements, whereas four-node thin shell-type elements with linear elastic behavior are
used to represent the upper sole. Vertical and horizontal steel reinforcements are
modeled using off-center bar-type elements. The concrete-steel adherence is
assumed to be perfect and the wall to be perfectly embedded at its bottom. The
meshwork used for the wall is presented in Figure 7.30.

The behavior of the wall is ruled by shear in the flanges as well as in the girder
web of the wall. As a matter of fact, the flanges change roles during the cyclic test,
as most of the shear imposed in the Y-direction is taken up again by the compressed
flange. During the 4 cm cycles, the steel reinforcements located at the flange-web
junction tends to buckle and the concrete starts to exhibit spalling. During the 8 cm
cycles, most of the steel at the ends of the flanges and the girder webs buckle, the
concrete gets crushed and collapse takes place because of the shear in the
compressed flange. The calculation-experiment comparison of the displacement-
load cyclic curves recorded during the USW3 test is presented in Figure 7.31. Figure
7.32 shows the drawing of the loads in both directions: when the displacement in
one direction changes, while keeping the displacement on the orthogonal direction
practically constant, the load required in the latter direction to keep the
corresponding displacement constant drops.

800
600
400
200
0
-200
-400

800

600

400

200

0

-200
-400

-600
-800

CALCULATION
EXPERIENCE

CALCULATION
EXPERIENCE

-1000
-100 -80 -60 -40 -20 0 20 40 60 80 100
Displacement at the top X (mm)

-600

-100 -80 -60 -40 -20 0 20 40 60 80 100
Displacement at the top Y (mm)

Figure 7.31. USW3 test (XY dir.): comparison with cyclic calculation

The effect of such coupling on the load-displacement diagram for both directions
then consists of quasi-vertical unloading, which can be observed in Figure 7.31.