Assessing the seismic vulnerability of masonry buildings - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 337

The FaMIVE method fi rst calculates the collapse load multiplier for each
possible mechanism of each façade in a building, given geometric and struc-
tural characteristics and constraints. To choose the most critical failure
mode leading to the greater possible loss (i.e. the most vulnerable condition
for the façade or building), the collapse multiplier for each failure mode is
weighted by the mass of walls and fl oors involved in the mechanism. The
characteristics of the identifi ed mechanism are then used to derive an
equivalent nonlinear single-degree-of-freedom (SDOF) capacity curve to
be compared to a seismic demand, and eventually to defi ne performance
points, as illustrated in the fl owchart in Fig. 13.2.
In Fig. 13.2,

I represents the collapse load multiplier for each possible
failure mode, and can be interpreted as the lateral acceleration capacity of
the structure. Hence fragility curves in terms of lateral acceleration capacity
can be directly derived. Alternatively, having defi ned an equivalent nonlin-
ear SDOF, and its representative capacity curves, four damage limit states
are computed, corresponding to damage limitation (DL), structural damage
(SD), near collapse (NC) and collapse (C), respectively. Details on the
derivation of the capacity curves and the computation of the limit states
are provided in Section 13.3.2 and 13.3.3.
In order to derive fragility curves, the next step consists of defi ning limit
state performance criteria to be correlated to damage states. This step
involves uncertainties, as very limited evidence exists to establish such cor-
relation over a wide range of building typologies and shaking levels. While
robust databases of damage states exist in the literature, no attempt has
been made so far to record permanent drift and corresponding ground
shaking in a consistent way, so as to provide empirical evidence for capacity
curves. As an alternative, a number of authors have worked on correlating
performance indicators and damage indicators to obtain capacity curves
experimentally, by way of shaking table tests or push-over test (e.g. Paquette
and Bruneau, 2006; Meisl et al. , 2007; Magenes et al. , 2010).
Paquette and Bruneau (2006) and Magenes et al. (2010) have performed
tests to defi ne in-plane behaviour of masonry house-models. Several issues
arise when extrapolating such data to building stock performance. The fi rst
problem is the scaling usually applied to specimens tested on shaking tables,
as full scale of more than two storey structures are too large to be tested
in most laboratories. The reduced scale invariably affects the strength of the
materials, and the stiffness of the overall structure. A second issue is the
limited representativeness of a particular masonry fabric (brickwork or
stonework, dressed or rubble) and the limited number of parameters studied
in any experimental investigations (such as slenderness of piers and layout
of openings). Most importantly the majority of those tests have been carried
out focusing only on the capacity of in-plane walls, while very limited
experimental work has been conducted on the characterisation of out-of-

λ