Probabilistic robustness capacity curves of R.C. structures in seismic zone

Catastrophic consequences, both in terms of human and material losses on high interest structures, caused by exceptional events, brought AEC sector experts to show growing interest towards structural safety in last decades. In particular, the new concept of structural robustness was defined: robustness is intended in a more specific sense as the capability to support extreme loads due to exceptional event trying to prevent disproportioned collapse. Several researches have already been done, and a lot of efforts are still spent by scientific community with the objective to better understand the problem and to propose solutions regarding the structural robustness theme. The probabilistic analysis approach, integrates in this continuously developing context trying to define a more accurate structural design process which implies increasing efficiency in the whole realization process (from the design to the construction), respecting the established safety thresholds and looking also at the principle of economic sustainability. The combination of these two issues, of paramount importance for structural engineers, lead to the adoption of a probabilistic analysis approach for the evaluation of the structural robustness and thus of the structural capacity of a building, with the final aim to define the reliability of a structure. Therefore, this master’s degree thesis represents another piece of this complex and articulated mosaic, that researchers are trying to build one tessera after other. In this work, probabilistic robustness analysis is performed on a 2D R.C. multistorey plane frame in seismic zone. After the preliminary design stage with the classical semi-probabilistic approach, a random sampling is conducted both on material properties and actions. The sampling was made using the Latin hypercube simulation (LHS) technique which is a particular Monte Carlo sampling method; therefore, 100 different structural models were generated from the random values of each sampled variable. Then, a displacement-control push-down analysis was conducted on each of the 100 models on Atena2D software, through a non-linear analysis, imposing a controlled displacement to the top of the central column of the ground floor which was removed. The resulting load-displacement capacity curves were thus exploited to determine the dynamic amplification factor (DAF), by using the energetic theory proposed by (Izzuddin, 2007). Successively 100 non-linear structural analysis were conducted simulating the instantaneous column removal and the subsequent dynamic amplification of loads by applying the DAF to the central spans of the frame; the obtained results lead to the probabilistic determination of the new capacity curves of the structural system. Results show that more or less in half of the cases, the 2D frame will ideally reach a condition of disproportioned collapse, defining therefore, a set of potentially critical situations for the structural system in case of extreme event, causing the loss of a column from the structure and the consequent disproportionate collapse.