Ricerc@Sapienza

In this paper, a simplified procedure for the evaluation of the seismic performance of bridge piers founded on caissons subjected to strong ground motions is outlined. To this end, the up-per-bound semi-empirical relationships proposed in [1] are considered for the estimation of the seismic performance, expressed in terms of the maximum and permanent values of the deck drift ratio attained during and at the end of the seismic event. These drifts were related to the period ratio Teq/T0 between the fundamental periods of the deck-pier-caisson-soil sys-tem and of the soil column in free-field conditions. The deck drift and the period ratios were extracted from the results of an extensive parametric study, where 14 different systems were subjected to 6 real high-intensity seismic records. In the parametric study, 3D dynamic anal-yses were performed with the Finite Element Method in the time domain, in terms of effective stresses but assuming undrained conditions and adopting an elastic-plastic constitutive model to reproduce the irreversible soil behaviour under cyclic loading. As 3D dynamic numerical analyses are not expected to become an every-day design tool, the period ratios Teq/T0 are evaluated through empirical and analytical relationships available in the literature as well and then compared with the ratios obtained from the parametric study, to assess the possibil-ity of using simplified relationships while still getting a reliable estimate of the deck drift ra-tio. It is shown that these relationships can be profitably adopted provided that a fair estimate of the equivalent shear wave velocity, depending on the intensity of the seismic inputs, is used.