L'Impiego di Modelli a Fibre per il Progetto d'Adeguamento Sismico di Telai in C.A. mediante Controventi Dissipativi

The reference methods for evaluation of performances of reinforced concrete frames subjected to strong seismic actions are the non linear static (push-over) analysis and the step –by step non linear dynamic analysis. The objectives of these analyses are the evaluation of plastic deformation demands in the structural elements that characterise the structure seismic behaviour. To this aims, robust and computational efficient models are required, that are able to capture the main detailing of the structure and element behaviour. Recently, the use of fiber elements for evaluation of seismic performance of reinforced concrete frame seems to be an attractive tool, due to the versatility in reproducing, with high level of accuracy, the cyclic behaviour of frame elements under seismic actions. In this field, the ineffectiveness of fiber model for prediction of ductility demand of elements characterized by cross sections having elasto-perflectly plastic or elasto plastic with softening moment-curvature relationship has been recently stressed (Spacone et al., 1996), due to localization of plastic deformation. However, when reinforced concrete and steel strain are limited to values of technical interest, the major part of reinforced concrete sections, even if loaded by high axial loads, exhibits a strain hardening behaviour. Thus localization effects are avoided, but loss of accuracy in the evaluation of the curvature demand can be found if the mesh of the fiber model has not been suitably chosen. In the paper, the first section is devoted to the investigation of mesh size effects in fiber models on accuracy in the evaluation of the relationship between displacement and bending curvature of reinforced concrete members. The analyses shown that it is opportune to calibrate the mesh size in relation to the reinforcement characteristics and the expected axial force level on the elements. In the second section these indications are employed for modelling reinforced concrete frames retrofitted by bracings with dissipative devices. Several criteria for the design of strength of the dissipative bracing are compared by non linear dynamic analyses of multistorey frames. The ineffectiveness of the bracing design criteria based on evaluation of global response parameters, such as roof displacement or interstory drifts, that lead to predict a large range of the strength of the dissipative device that optimise the bracing system performance is shown. The analyses prove that the behaviour of frames with collapse condition dominated by available ductility of base column plastic hinges is strongly influenced by the strength of the dissipative devices. Design criteria based on control of global response parameters are not able to predict the reduction of bracing system performance due to reduction of available ductility due to high axial stresses induced by the bracing system. The reliability of an optimising criterion for bracing strength design, controlling local and global damage indexes, are shown.