Influence of modelling assumptions in the expected loss evaluation of a precast industrial building

The economic losses arisen in the industrial sector after the Emilia earthquakes in 2012 highlighted the importance of conducting reliable seismic assessment analyses on the existing industrial building stock in order to ascertain both safety and potential losses associated to seismic events. To accomplish to such task, an accurate representation and quantification of the actual vulnerabilities in such buildings is required and reliable structural models need to be adopted. The paper investigates how various assumptions and levels of sophistication in finite element modelling affect the results in terms of economic losses, herein assumed as the reference decision variable. After the definition of the main seismic vulnerabilities of precast industrial buildings typical of the Italian territory, different types of finite element models are adopted and non-linear time history analyses conducted; in particular, the reinforced concrete fork at the top of the column is modelled in different ways, also considering the seismic retrofit. Appropriate fragility curves under selected engineering demand parameters are defined and provided within the Performance Based Earthquake Engineering methodology developed at the Pacific Earthquake Engineering Research Center for the assessment of the expected losses under a scenario-based earthquake. The influence of the modelling assumptions in the seismic risk estimate is evaluated. The results indicate that for the considered case study and in the absence of loss of support, simplified single-column models are suitable to estimate the expected losses.