Response of the Fatih Sultan Mehmet Suspension Bridge under spatially varying multi-point earthquake excitations

The study aims at investigating the structural behavior of the Fatih Sultan Mehmet Suspension Bridge, i.e. the second Bosphorus Bridge in Turkey, under multi-point earthquake excitations, and determining the earthquake performance of the bridge based on the results obtained from this analysis. For this objective, spatially varying ground motions in triple direction were produced for each support of the bridge considering the Mw=7.4 scenario earthquakes on the main Marmara Fault. In order to simulate the ground motions, modified stochastic finite-fault technique was utilized. Taking the ground motions into account, non-linear time-history analysis was carried out, and the results obtained from the analysis were compared to those from uniform support earthquake excitation to identify the effects of multi-point earthquake excitations on the seismic performance of the bridge. From the analysis, it was determined that modal response of the towers and the deck was mostly effective on dynamic response of the entire bridge rather than other structural elements, such as cable and approach viaduct. Compared to the results obtained from simple-point earthquake excitation, noticeable axial force increase in the cable elements was obtained under multi-point earthquake excitation. The changes at the main cable and the side span cable were determined as 21% and 18%, respectively. This much increase in the cable elements led to increase in axial force at the towers and in shear force at the base section of the tower column. These changes in the structural elements were closely related to response of the deck and the towers since they had considerable contribution to response of the entire bridge. Based on the findings from the study, spatially varying ground motions has to be considered for long span suspension bridges, and the multi-support earthquake analysis should be carried out for better understanding and obtaining reliable results necessary for retrofitting and performance evaluation. (C) 2016 Elsevier Ltd. All rights reserved.