Models for Simulating the Seismic Response of Concrete Structures - Dynamic Behavior of Concrete and Seismic Engineering - page 270

models are suited to the phenomena to emphasize (material models with damage and
models for the interfaces that give rise to strong local non-linearity). Such models
are especially recommended to access global and collapse-mode behaviors.
A good calculation experiment correlation should not leave aside the difficulty in
finding local results (like crack opening or steel deformation) for important damage
levels. Such difficulties are all the more awkward as the criteria that herald damage
of the structure (and therefore its potential collapse) are generally local in para-
seismic regulations.
To obtain more realistic apprehension of local values, a solution lies in more
complete 2D and 3D discretization. We could state that the plastic model coupled
with a correct description of the limit conditions and of loading allows us to justify
the essential mechanisms of the behavior of a wall under cyclic loading. The main
drawback of this local approach lies in the amount of calculation generated by the
great number of degrees of freedom used, far more than with simplified models.
Nevertheless, the plane stressed walls are quite suitable for 2D modeling, and from
this point of view, current computer means are quite well adapted. The behavior of
walls that have a U-shaped section and are loaded in bi-bending is modeled using a
multi-layer thin shell 3D approach, based on the 2D local model. In the last case,
although the amount of calculation generated is higher than the 2D approach, it is
still quite reasonable and much lower than massive 3D modeling.
Whilst highlighting these model's possibilities, we should not conceal their
limitations and it is important to note that the local phenomena that take place at the
concrete-steel interface can only be represented approximately according to the
concrete model used. The shear transfer factor K allows some of the transmission of
the shear stresses in the cracking plane to be taken into account, but it cannot
correctly represent the mode II cyclic energy dissipation.
Furthermore, for all the presented models, the modeling quality is affected by the
parameter identification procedures used. Some experimental procedure regulations
for the commonly admitted characterization of materials could prove quite useful.
7.6. Bibliography
[CAZ03]CAZAUX G., Modélisation tridimensionnelle du comportement sismique
d'ouvrages en béton armé - Développement de méthodes simplifiées, PhD Thesis, LMT-
ENS de Cachan, 2003.
[COI 98] COIN A. et al. , “Recherche CAMUS, Rapport final, Présentation générale,
Conclusions et perspectives”, Ministère de l'Equipement, Plan Génie Civil, Convention
9670011 , 10 pages, 1998.