Seismic Vulnerability of the Highway System

Task C3-1: Steel Truss Bridge Braced Pier and Substructure Connection Behavior

Subject Area:  Special Bridges - Substructures and Superstructures 
Research Year 2

Principal Investigator(s) and Institution(s)

Michel Bruneau and John Mander, University at Buffalo


A large number of steel truss bridges have been built throughout the Untied States, many in zones of moderate to high seismicity. These are often supported by truss braced piers. These piers have typically been designed to resist wind forces, but not earthquakes. Structural analyses of such bridges often reveal that many key structural members along the load path followed by the seismically induced forces may buckle or suffer brittle fracture of their non-ductile connections. However, seismic evaluation remains difficult due to the lack of knowledge on the cyclic inelastic behavior of built-up members of the type typically found in these bridges, and on their riveted connections. The objective of this task is to analytically and experimentally develop and improve knowledge concerning the seismic behavior of steel truss substructures.

Although some testing of bridge-specific components has been conducted by other researchers, results to date have mostly been project specific; as a result, it is difficult to draw general conclusions from them. This task will therefore attempt to provide theories and results that can be broadly applicable to many steel truss bridges which share similar structural characteristics and details.

Research Year 1 Progress

During the first year of this task, an extensive survey was conducted to investigate those commonly used shapes and details that have been historically used for built-up members used in the braced towers of bridges. This survey showed that braces almost universally consist of latticed built-up members with riveted connections. Several common cross-sectional shapes have been identified. Evidently, these shapes evolved depending on the degree of compression loads these members were expected to carry. There are essentially four potential failure modes that consist of either global or local buckling that are either in or out of the plane of the brace. To this end, based on representative details, an experimental program has been devised whereby several cross-sectional shapes and geometric configurations of X-bracing will be explored to ascertain the inelastic deformation capability of these critical substructural elements. Work has already commenced on this experimental investigation in two ways. First, the experimental infrastructure for the experiments has been prepared; and secondly, the materials to fabricate the specimens have been ordered. Due to limitations of the initial year funding only one or two specimens will be fabricated for testing in Year 1. Fabrication is quite costly due to the extensive number of rivet-like connections that must be made. 

During the design phase of this experimental program, companion analysis work has also been conducted. First, the yield (onset of damage) limit state deformation capacity has been identified. This study shows that substructure drifts in the order of only 0.25% can be expected prior to causing permanent damage to the steel X-bracing. The degree of post-yield drift capacity will be ascertained as one of the primary products of the Year 2 experimental program which is described in what follows.

Research Year 2 Tasks

As described above, an experimental investigation will be conducted to identify failure modes and displacement limit states for a selection of critical members used in X-braced steel truss bridge substructures. A companion study to this experimental program will involve modeling the results in two ways. First, simple strength and deformation models based upon rational mechanics will be devised that predict strength and displacement limit states, principally for monotonic behavior. Such simple models are necessary for use in either pushover and/or plastic analyses. A second class of model that will build upon this information is to formulate hysteretic rules that will model the inelastic cyclic behavior of built-up braced members and their riveted connections. These models will be used in well-known nonlinear inelastic time history analysis computer programs such as DRAIN or IDARC to investigate the seismic resistance of entire steel truss bridges. Knowledge from this experimental and analytical study will be used to inform a separate project on the development of retrofit measures for those members that are seismically vulnerable.


Behavior models and analysis guidelines for truss bridge braced piers and substructure connections.

Technical Challenges

The development of reduced scale specimens with all the intricate details that will faithfully reproduce the expected performance of prototype built-up bricked members.


return to Task Statement List