The 14 th World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China SEISMIC RESPONSE OF BRIDGES WITH INTEGRAL DECK-ABUTMENTS: AMBIENT VIBRATION TESTING J.C. Carvajal 1 , C.E. Ventura 2 , S. Huffman 3 and F. Yao 4 1 Research Assistant, 2 Professor, 4 Lab. Manager, Dept. of Civil Engineering, University of British Columbia, Vancouver, Canada 3 Bridge Seismic Engineer, Ministry of Transportation and Infrastructure - TRAN, Victoria, Canada Email: jccu@civil.ubc.ca, ventura@civil.ubc.ca, Sharlie.Huffman@gov.bc.ca ABSTRACT: Integral abutment bridges are single or multiple span structures consisting of a continuous concrete deck integrated with abutments and a flexible support system. These structures have become very popular due to the elimination of costly and maintenance prone expansion joints and bearings. However, the exclusion of joints causes the superstructure to interact more closely with the soil embankments, making the evaluation of vibration properties and seismic demands a process that involves many uncertainties. The purpose of this paper is to contribute to the better understanding of this type of bridges by investigating in detail their vibration characteristics at low levels of excitation. The paper presents the identification of modal vibration properties of six bridges with integral deck-abutments using ambient vibration tests. Results indicate that the motion of the embankments causes the superstructures of some single-span bridges to deflect with an operational mode similar to the first vertical natural mode of vibration. This behavior suggests that the seismic response of the bridges in the vertical direction may be strongly influenced by the seismic response of the embankments and the abutments. Steel bridges have developed thick cracks at the end of the approach slabs in the embankments. These bridges have the lowest fundamental frequencies of all the tested bridges. The presence of these cracks suggests that the abutment foundation may be too flexible for this type of superstructures. The identification of predominant frequencies of vibration of the embankments is done with a proposed technique named “cross response spectrum”. This technique is useful when the identification of predominant frequencies can not be easily determined by more traditional identification techniques such as the power spectral density and the H/V ratio. KEYWORDS: natural frequency, mode shape, modal damping, response spectra. 1. INTRODUCTION Seismic assessment of bridges requires an educated prediction of their seismic capacity and demand under various earthquake scenarios. In such an evaluation, there are many sources of uncertainty that engineers have to deal with. These uncertainties are due to: a) the selected loads and material properties, and b) the prediction of the structural response. One of the sources of uncertainty in predicting the structural response is the soil- foundation-structure interaction (SFSI). SFSI can significantly change the response of the structure. For instance, the foundation response, which determines the input to the structure, can differ from that of the site, which in many cases is assumed as the input to the bridge. Another source of uncertainty in the structural response is the soil-abutment-structure interaction (SASI). SASI is usually modeled by replacing the backfill material with equivalent springs and dampers connected to the wall abutments. This approach simplifies the problem; however, it ignores the time-dependent variation of inertial forces induced by the backfill against the abutments during a seismic event. SFSI and SASI can significantly change the actual vibration properties of the bridge in comparison to the predicted with the commonly fixed-base assumption. In this regard, The University of British Columbia (UBC) and the Ministry of Transportation and Infrastructure (TRAN) initiated in 2007 a joint study to evaluate the seismic response of bridges with integral deck-abutments in the province. The study employs 3D finite element modeling of the structures and their soil embankments to evaluate the effects of SFSI and SASI on these types of bridges for different scenarios of seismic hazard. The first