A comparison and evaluation of different soil-structure-interaction approaches for bridges

Abstract

During the last fifty years, bridge construction has increased extensively throughout the world, including on areas with bad soil conditions, to meet the transportation needs of expanding urban areas. Although Soil-Structure-Interaction (SSI) procedures for performance-based design of buildings have been introduced in design guidelines, seismic provisions are not clearly stated for bridges. There are two main approaches to include SSI in performance-based design of the bridges; direct method and substructure method. In the direct method, bridge and soil systems are analyzed as a single system under seismic shaking, defined at bedrock. As an alternative, the substructure method is introduced to solve the system as substructures in two stages, called kinematic interaction and inertial interaction. The nonlinear response of piled foundation systems of bridges are subjected to kinematic interaction; whereas, the nonlinear response of superstructure is subjected to inertial interaction. In this study, first, he linear design of two different bridges are introduced by considering their geometry and the number of spans. Bridge-I has three spans with uneven pier heights, and Bridge-II has four spans with identical piers and the geometry. Both bridges are designed based on a response spectrum created according to site response analysis and used in the performance-based design of the bridges. 19 different records are selected and scaled according to the criteria given in AASHTO LRFD Bridge Design Specifications Article 4.7.4.3.4 (AASHTO, 2012) Seismic Design Guidelines. Seismic records are categorized with respect to soil parameters, chosen for both strength-based and performancebased design of bridges. Secondly, using the direct method, the Nonlinear Time History (NTH) analyses are performed for both bridges to investigate the behavior of structural elements. The nonlinear responses of the bridges are re-calculated by using the substructure method, including the kinematic and inertial interactions. Responses of the structural elements are combined according to commonly-used combination rules. Finally, results of these methods are compared with each other, as well as the linear response of the structures, to underline how the behavior of the structures vary according to different analysis methods.