State of the art in seismic vulnerability

Abstract

When performing vulnerability analysis and in absence of statistical data about vulnerability of some types of widespread buildings, unique structures, historical monuments etc, the development of numerical methods and approaches of evaluation and identification of vulnerability grade is shown as an important tool for different purposes connected to risk assessment from natural disasters.

For adequate vulnerability assessment the design model of a building should take into consideration the soil-structure interaction, spatial effects and real structural features including existing damages.

Probabilistic (empirical and judgment vulnerability methods) are the simplest but they need huge damage database from old earthquakes with different size and in different regions and they are valid if they are applied in locations with the same building characteristic as the database. Anyway it is not so clear if the damage from moderate earthquake can be extrapolated to higher earthquakes.

Currently the most recent vulnerability method applied in loss estimation studies (HAZUS approach) involves the use of numerical methods (deterministic) by means of the Capacity Spectrum Method and the damage state of a building is taken to be defined by the interstorey drift ratio at the most deformed level of the building. This parameter has been confirmed as having the best correlation with structural damage. This method has its power in the use of as much as possible information on soil conditions and building stock, but in those cases in which this information is scarce, a simplified approach (with less computational complexity), as demonstrated by Pinho et al. (2002) looks to be valid.

Anyway, these deterministic (predictive, theoretical or numerical) methods have to be calibrated with damage data from old earthquakes if possible and future earthquakes once they have happened.