Incremental dynamic analysis (IDA) applied to seismic risk assessment of bridges - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 558

2008). For IDA the in-cycle strength degradation is very important for
prediction of the collapse capacity and the global dynamic instability of the
structures, and is often considered by the negative post-capping stiffness
defi ned in the component backbone curves (see Fig. 21.2).
Modern bridges are designed and detailed to meet the seismic code
requirements for ductile response so that the fl exural response governs the
response and the collapse mechanism of the bridge. When structures with
poor detailing are studied or when other collapse modes are possible,
degrading collapse modes, such as shear failures or shear-fl exure failures,
can be considered directly in the structural models. Including code-
conforming confi nement reinforcement in the plastic hinge region of bridge
columns will prevent such undesirable failure modes. Where such collapse
modes cannot be incorporated directly in the models, often they are con-
sidered indirectly by imposing some drift limits corresponding to such col-
lapse modes (i.e., non-simulated component limit state criteria). That is,
after performing IDA, the collapse capacities can be modifi ed for other
collapse modes such that the corresponding IM values to the non-simulated
collapse modes (in terms of DM) will be considered as the collapse capacity
of the structure. However, this simplifi cation in modelling through applica-
tion of the non-simulated component limit state criteria should be accounted
for by increasing the modelling uncertainty in the overall seismic per-
formance assessment of the structure. The model proposed by Elwood
(2004) is an example of the models that can predict the drift corresponding
to shear failure and axial failure of columns with insuffi cient transverse
reinforcement.
The infl uence of the abutments on the seismic response of the bridges
can be included in structural modelling. However, the effects of abutments
may be conservatively neglected if they do not signifi cantly infl uence the
seismic response. An alternative approach may be to analyse the bridges
for two cases of restrained and unrestrained movements at the abutments
to evaluate the seismic response of the structure and to identify the control-
ling case. For example, in the transverse response shear keys are often used
to restrain the transverse movements of the superstructure at the abut-
ments. The shear keys may either be designed based on the capacity design
concepts so that they will survive under high ground motion intensities or
they can be designed as fuses to fail at low ground motion intensity levels
to prevent damage in the abutments and piles. Therefore, depending on the
design philosophy for such elements, different modelling approaches may
be accepted. The most direct approach would be to include the degrading
behaviour and failure of shear keys in the structural models. The shear key
models by Megally et al. (2003) can provide the required information for
this purpose. The spring abutment model developed by Aviram et al. (2008a)
can be used when the effects of abutments are directly considered in