Seismic fragility matrices for large scale probabilistic structural safety assessment

Structural safety assessment of urbanised areas exposed to seismic risk is a complex task and can be approached at different scales. Fragility functions are the most effective tool to perform an estimation of the structural risk, since they describe the exceeding probability of the limit state of interest given a level of ground shaking. The formulation of a fragility function is based on the selection of an Intensity Measure of the ground shaking, commonly represented by peak-ground or spectral parameters or by macroseismic intensity grades. In the present paper, a novel approach to estimate the seismic vulnerability at a large scale of urbanised areas is presented. The scenario earthquakes are selected according to the current formulation of the seismic hazard disaggregation provided by National and International codes, i.e. through the disaggregation in terms of magnitude (M), epicentral distance (R) and number of standard deviation from the median ground motion (epsilon). To this aim, a vectorial seismic Intensity Measure is assumed to compute structural safety at large scale, and a change of variables is made to turn the traditional fragility curve into a matrix. Typological fragility matrices are processed and combined to define the so-called Urban Fragility Matrices, representative of the vulnerability of selected urban sectors. The procedure is then implemented and validated referring to the city of Pozzuoli (Southern Italy) and its effectiveness validated with reference to the damage survey relative to the 4th October 1983 seismic event which struck the city, resulting a good matching between predicted and observed scenarios.

Automated summary

This paper presents a novel approach to estimate the seismic vulnerability of large-scale urban areas. The method involves selecting an appropriate intensity measure, using matrix transformation to calculate structural safety, and validating the effectiveness with a specific case study.