Inelastic response of RC frames under different modeling assumptions.

In recent years the approach to design of structures in seismic areas has been undergoing a substantial evolution, with the development of the so-called Performance Based Seismic Design (PBSD). In this respect, an essential event has been the preparation of the Vision 2000 [1] document by the Structural Engineers Association of California, which has led to definition of ‘seismic performance objectives’ as the ‘coupling of expected performance levels with expected levels of seismic ground motion’, thus extending the wellestablished concept of designing structures to resist minor earthquakes with no damage, moderate earthquakes with no structural damage and severe earthquakes without collapse. Performance levels are usually defined as a function of deformation parameters; for framed structures, response parameters universally considered are interstorey drifts and plastic rotations. It is clear that one fundamental problem is quantifying values of response parameter limits corresponding to the different performance levels. Another crucial problem is the adoption of refined computational models, in order to be able to predict values of the assumed response parameters with accuracy. This paper deals with the latter issue, with reference to RC framed structures. In this field, a number of computational models has been developed in the past, ranging from simplified global models to very refined local models, as evidenced in [2]. The former ones are not adequate for PBSD, the latter ones can be prohibitively costly in large dynamic simulations. Therefore, it is believed that an intermediate model, such as the one presented in [3, 4, 5], based on the decomposition of frame members into different sub-elements, can be really suitable for carrying out PBSD procedures in a reliable manner. In particular, one of the main feature of this numerical model is its ability to include in the simulations joint flexibility due to bar pull-out at the beam-column and column-foundation interface. Herein, effects of such enhanced modelling are evaluated in a PBSD format, by comparing the predictions of interstorey drifts and plastic rotations with the limits provided by FEMA 356 [6]. The performance analyses are conducted on a three bay four story RC frame designed according to Eurocode 8 (EC8) [7] and subjected to two different earthquake records. The obtained results demonstrate that inclusion of bar pull-out influences significantly the two considered response parameters, regarding both values and distribution throughout frame members.