|Abstract:||The American Association of State Highway and Transportation Officials (AASHTO) seismic provisions were updated in 2008. These provisions increased the design earthquake return period from 500 to 1000 years, eventually demanding special structural systems to counteract seismic forces and therefore increasing construction costs. The updates were based on the seismic practices in the western United States, which focus on bridge configurations that dissipate energy either by plastic deformation of the piers or by specially designed isolation devices placed between the superstructure and substructure.
The Illinois Department of Transportation (IDOT) has focused on this second isolation approach, however considering only conventional bearings, in order to formulate a cost-effective “quasi-isolation” concept targeted to the seismic hazard and typical bridge configurations in Illinois. An Earthquake Resisting System (ERS) based on quasi-isolation uses elastomeric bearings and is different from conventional seismic design by permitting bearing anchorages to fracture during a design earthquake. The ERS also depends on subsequent bearing deformation and sliding to accommodate seismic demands.
IDOT has organized two research phases to analyze and develop the concept of an ERS based on quasi-isolation. Phase I performed an experimental investigation of the seismic response of bridge bearings and formulated structural models of typical bridges based on information obtained after testing these structural components. The structural models of this phase only considered regular bridges and a simplified abutment model. The structural performance of these models was assessed for various hazard levels by using a suite of synthetic ground motions based on representative seismic records of the New Madrid Seismic Zone (NMSZ), a geographical region that encompasses several locations with high seismicity, including southern Illinois. Phase II continued the assessment process of the proposed ERS by formulating structural models that considered additional features such as skew angles and a detailed abutment model that included elements such as approach slabs and wingwalls. In this phase, a suite of synthetic ground motions specifically formulated for Cairo, Illinois, the geographical location with the highest seismic hazard of the state, was developed to analyze the structural performance of these models for design hazard levels.
The principal objective of the present work is to evaluate modeling sensitivity between Phase I and Phase II models in order to determine to what extent the additional elements included in Phase II models affect overall structural response. In order to perform a consistent comparison, bridge configurations that can be found in both Phase I and Phase II parametric studies were selected. The comparison encompassed static pushover analyses and nonlinear dynamic analyses. Static pushover analyses were considered to determine similarities and/or differences of bridge response characteristics such as force distribution among substructures, sequence of limit state occurrences, fusing of sacrificial connections, and vulnerability of critical bridge components. Nonlinear dynamic analyses were considered to assess the seismic performance of Phase I and Phase II models based on a comparison of the number of occurrences of various limit states. The synthetic ground motions developed during Phase II were used for this comparison.
This study found that additional Phase II elements, such as approach slabs, increased structural stiffness at abutments in the longitudinal direction of analysis, which allowed a redistribution of forces at this location. This redistribution precluded the concentration of forces in sacrificial connections and reduced superstructure displacements. In general, these conditions diminished the occurrence of fusing limit states at abutments and damaging limit states at intermediate substructures in comparison to Phase I models.
Likewise, in the transverse direction of analysis, it was observed that additional elements, especially wingwalls, increased abutment stiffness that allowed a redistribution of forces that essentially reduced the concentration of demands on sacrificial connections at that location. Generally, the stiffness increase observed in the transverse direction was lower compared to the longitudinal direction.
Finally, according to the results of nonlinear static analyses and nonlinear dynamic analyses, it was not possible to identify a marked difference in the occurrence of limit states between Phase I and Phase II models, especially regarding damaging limit states. Even though the formulation of Phase II models resemble more closely the real structural configuration of seat-type abutment bridges, the Phase I models require less computational resources to be analyzed. For this reason, it is possible to employ the more simplified Phase I models as a preliminary assessment tool, before using more complex formulations (as needed) for definitive structural analysis of seat-type abutment bridges.