ABSTRACT
The earthquake-induced deck-abutment interaction often triggers drastic change of the effective mechanical system of bridge structures. The key to a numerical study of deck-abutment interaction is the simulation of the contact phenomenon. The present paper compares two simulation approaches: a compliance (or gap element) approach and a nonsmooth dynamics approach. The two approaches are assessed with respect to their ability to predict the measured response of a straight, 4-span bridge model in an experimental shake-table study. The paper also investigates the sensitivity of the deck response to the contact element stiffness and the rotation of the abutment. The results show that the deck rotation predicted by these two approaches differs notably.
