Highway bridges are designed to resist most “normal” loads elastically. For example, steel members are not expected to yield and no permanent deflections result due to the design dead, live or wind loads. On the other hand, seismic demands can be so large that bridges cannot resist these forces elastically. Yield stresses are exceeded and permanent deformations are likely to occur due to earthquake loading.
Seismic design of common bridges using the AASHTO LRFD Bridge Design Specifications and the AASHTO Guide Specification for LRFD Seismic Bridge Design methodologies rely on an elastic analysis from which inelastic demands are estimated. Seismic bridge response is classified in four Seismic Design Categories (SDC). SDC A is the lowest seismic hazard and presumes an elastic response during the design seismic event. Bridges located in SDC B, C, and D are likely to experience inelastic demands during the design seismic event. About half of the country is subject to SDC B, C, or D seismic demands. The equal displacement assumption has the same consequences on bridges designed using the force-based AASHTO LRFD Bridge Design Specifications and the displacement-based AASHTO Guide Specifications for LRFD Seismic Bridge Design. This is not just a “seismic state” concern.
It is difficult to perform nonlinear analysis of bridges. Inelastic analysis requires special computer programs that can calculate bridge response due to ground shaking (i.e., accelerograms that are themselves difficult to develop). As with most sophisticated analysis, user knowledge and experience is required to develop, interpret, and verify the computer results. Due to these challenges, engineers (and the AASHTO specifications) use an approximate method of seismic analysis wherein the inelastic bridge response is estimated using an elastic bridge model. These elastic analysis methods are founded on the “equal displacement assumption” in which the “real” inelastic bridge is replaced by a fictitious elastic bridge that never yields.
The equal displacement approximation was introduced over 60 years ago, but is not based upon significant physical validation. Advances in numerical and physical methods suggest that the equal displacement approximation underestimates the “real” seismic displacement demands. Consequently, bridges may not perform as intended when using elastic analysis methods used for most designs. The potential underestimation associated with the AASHTO provisions was verified by the recently completed NCHRP Project 12-106.
The goal of this project is to examine the equal displacement approximation and, if necessary, develop simple, reliable adjustment factors that can modify the elastic displacement predictions to better match the “real” inelastic displacement response. These design recommendations will allow engineers to better predict seismic demands, resulting in safer, more reliable bridges.
The objective of the research is to develop robust yet simple procedures for calculation of inelastic displacement demands in bridges subject to earthquake demands.