Derivation of fragility curves to assess and compare the effectiveness of retrofitting strategies in URM buildings

Recent earthquakes have highlighted the seismic vulnerability of existing masonry buildings, pointing out the urgent need to design high-performance and non-invasive strengthening interventions. To this aim, reliable numerical models and assessment procedures are necessary to evaluate their effectiveness and quantify the effects on the structural seismic performance of the building. In this context, the article describes a procedure for the seismic assessment of existing masonry buildings and the design of retrofitting strategies. The procedure is based on a suitable nonlinear numerical model and consists of three phases. In the first phase, the numerical model is defined. In the second phase, the seismic assessment of the building in the as-built configuration is performed, and the required strengthening interventions are designed using nonlinear static analyses. In the third phase, the actual effects of the proposed interventions are evaluated through nonlinear dynamic analyses and the derivation of fragility curves. To test its reliability, the article describes the application of the procedure to two case studies representative of ordinary unreinforced masonry buildings placed in Italy and considered both in the as-built and retrofitted configurations. Their seismic performance was assessed through several nonlinear dynamic analyses performed according to the multiple-stripe approach and employing an equivalent frame modeling strategy. Different strengthening techniques were investigated, designed to conform with the Italian Technical Code, and modulated to achieve a safety index higher than 1 or 0.8 (allowed by the code for existing buildings) as a function of two different site categories (stiff and soft soils). The resulting fragility curves were used to quantify the improvement of the seismic performance guaranteed by the different strengthening strategies and compare their effectiveness concerning two performance conditions (namely usability-preventing damage and global collapse limit states) to identify the best solution for the two considered site categories.