Seismic Assessment of Evin-Valley Bridge by 3-D Inelastic Dynamic Analysis

Document Type : Seismology and Engineering Seismology

Authors

K.N. Toosi University of Technology

Abstract

Major damage was observed mostly in the older bridge structures in the Northridge (1994), Kobe (1995) and Taiwan (1999) earthquakes. The most extensive damage included flexural/shear failure of substructure members and superstructure unseating at simple supports or expansion joints. In general, similar types of damages demonstrate a similar nature in view of the seismic behavior of older bridge design in all three earthquakes. As modern bridges have not been significantly affected during the recent ground motions, no reliable judgment could be made on the seismic performance of modern bridges structures.The seismic performance of Evin-Valley Bridge, a newly built slab-on-girder bridge is investigated analytically at the damage control limit state. A 3-D model of the bridge was built in DRAIN 3DX computer programme using a fiber-section beam-column element to represent inelastic behavior of RC substructure members. Elastomeric bearing pads, shear keys, expansion joints and the abutment backwalls were included in modeling of cyclic behavior of each component. The superstructure girder-beams were assumed to remain elastic and compressive elastic springs were used to represent the soil effect. A free vibration analysis of the multiple-part structure, assuming open gaps and expansion joints, showed the combined influence of a broad number of modes of vibration in dynamic response of the bridge. However, since the gaps are predicted to frequently close and reopen under earthquake forces, such results should not be relied on predicting of the seismic response. The results obtained from dynamic analyses using the Naghan (1977), the Northridge (1994) and the Kobe (1995) acceleration records show that the seismic demand values in the substructure elements are much less than the existing member capacities. The results also indicate that the dynamic response values are not comparable with the earthquake demands obtained from the equivalent static method and a large difference is observed between the results of two methods.

Keywords