Reliability of pushover analysis - Based methods for seismic assessment of R/C buildings

Alternative assessment procedures based on the pushover concept are evaluated, including the ATC-40 method based on the so called "capacity spectrum", and a recently proposed method where the seismic demand is expressed in the form of inelastic response spectra compatible with the level of inelasticity expected in the structure assessed. The type of structure studied is a multistorey reinforced concrete (R/C) frame designed to Eurocode 8 for different ductility classes ('Low', 'Medium', and the new class 'E' introduced in the final version of the code). Three different models are considered for the frame members, i.e. trilinear, fully cracked bilinear, and bilinear according to ATC-40. The inelastic spectra - based procedure is applied using two alternative proposals for deriving inelastic spectra. The ATC-40 method is applied assuming two different classifications for the frames that result in different equivalent damping; furthermore, for each type assumed, damping is calculated using either the ATC or the Eurocode 8 relationships. The foregoing lead to a total of 6 different methods applied to each of the three structures modelled in three different ways, hence a total of 54 assessments. Subsequently, the outcome of the methods based on pushover analysis is assessed in the light of inelastic time-history analyses of the frames for a total of five accelerograms. Results from the foregoing wide range of analyses indicate a number of significant trends, such as the significant effect of the initial stiffness assumption, the inappropriateness of assuming Type A damping for these structures, although they clearly fall within the "good behaviour" category specified in ATC-40, and the satisfactory performance of all Eurocode designs.