A RELIABILITY COMPARISON BETWEEN DIFFERENT ROBUSTNESS STRATEGIES IN SEISMIC ZONE

The theory of structural robustness has only recently emerged in the field of structural civil engineering, as a consequence of both minor and major collapses, one of which being the World Trade Centre terrorist attack (New York, 11th September 2001). Normally, buildings are designed to withstand normal actions, whose types and intensities are defined according to the building’s importance. In reality, structures are not always able to withstand exceptional events, such as impacts, explosions, and fires; these scenarios have a low probability of occurring, but can have catastrophic results in terms of loss of life, environmental and economic damage, as well as disruptions to public utility services. By definition, structural robustness is an intrinsic characteristic of a structure and indicates its ability to avoid disproportionate or progressive collapses as a result of local damage. A structure is thus considered to be robust when, after the loss of one or more sections, it is able to establish other load paths and properly redistribute the forces to the remaining portions. In this work of thesis, a probabilistic method is used to evaluate the robustness of a reinforced concrete building. This is done because a reliability evaluation of the issue, taking into consideration the uncertainties, may be useful talking about robustness analysis. Two different frames are analysed, designed in an area of high seismicity according to the directives of the Italian and European regulations. The former has been designed according to code rules while the latter has been improved following other prescriptions, coming from previous experimental studies, to enhance structural robustness. One hundred different samples of the materials and load characteristics for each frame were taken into account applying a probabilistic approach, the Latin Hypercube Sampling (LHS). A FEM software, called ATENA 2D, has been used. The main analyses that have been performed, for all the one-hundred scenarios for each frame, are of two types: a pushdown analysis and a reliability analysis. The former was necessary to compute the dynamic amplification coefficients, point where the curves representing the internal energy and the external work meet, to be applied to the loads in the second type of analysis, after the column is removed. The latter is needed for the reliability evaluation of the structural robustness. These dynamic amplification coefficients, which varied for each of the one hundred situations, were analysed using a technique given by Izzuddin, which involves performing a pushdown analysis. The resulting curves are the so-called load-displacement capacity curves, through which it was possible to become aware of the best behaviours in the different cases treated. After that, the reliability analysis was carried out for each of the one hundred different scenarios for each frame, which made it possible to validate the procedure that Izzuddin had proposed; all of the scenarios in which the equilibrium could not be reached did not effectively meet convergence criteria in the probabilistic analysis, which confirmed the results of the initial phase. In the end, the strains at different points of frames’ sections were monitored to determine the local probability of failure. This was done for both the concrete and the reinforcing components of the structures. As a direct consequence of this, a global reliability evaluation and a critical comparison between two different designing criteria were carried out.