Assessing the seismic vulnerability of masonry buildings - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 328

wherein scores are attributed to each signifi cant vulnerability component
(UNDP/UNIDO, 1985). Several of these methods have been developed in
the past 30 years and the more robust and commonly used are reviewed in
Section 13.2.1.
Recent dramatic increases in computational powers and consequent
improvement and refi nement in numerical modelling of relatively complex
structures, using either static or dynamic approaches (e.g. FEMA-356, 2000;
Vamvatsikos and Cornell, 2005), has facilitated greater exploitation of ana-
lytical approaches for vulnerability assessment. Analytical methods need
experimental validation of the various parameters used to defi ne the vul-
nerability, and so far relatively little experimental work on the behaviour
of masonry structures has been undertaken by the international community,
albeit a wide variety of unreinforced masonry typologies, both historic and
modern, exists. The framework within which analytical methods can be
applied and their limitations are presented in Section 13.2.2.
An attempt to exploit positive aspects of the two classes of methods
above is made with the so-called hybrid methods. These combine numerical
input/output from analytical models with statistical and probabilistic data
to defi ne exposure and vulnerability distribution. This allows the analytical
burden to be reduced while grounding results in a geographical context.
Applicability and reliability of such approaches are constrained by the
ability to defi ne certain mechanical and structural characteristics in numeri-
cal terms and by the need to defi ne a common approach for treating the
various sources of uncertainty associated with vulnerability, exposure and
hazard. This is further discussed in Section 13.2.3.
The review and discussion of existing methods sets out the rationale for
a combined method, called FaMIVE (Failure Mechanism Identifi cation
and Vulnerability Evaluation), and developed by the author and her co-
researchers, in the past decade. The hypotheses and analytical approach
to derive capacity curves and fragility functions are presented in detail in
Section 13.3. Results obtained from the FaMIVE method are presented in
Section 13.3.2. in terms of damage limit states and drift, and compared with
the recommendation of Eurocode 8 and the experimental evidence to
discuss issues of validation and calibration. The relevance of different cat-
egories of input parameters is discussed with respect to mean capacity
curves for different cases. A method for deriving performance points in
the acceleration/displacement response spectra plane is introduced and
results are evaluated in terms of Eurocode 8 recommendations on knowl-
edge level and confi dence factors in Section 13.3.3. Finally, fragility curves
are derived for various cases (e.g. different locations and limit states). A
discussion on treatment of various sources of uncertainties is presented in
Section 13.4 and conclusions and future research needs are summarised in
Section 13.5.