Tools for Gusset Plate Evaluation

The collapse of the I-35W Bridge in Minneapolis, Minnesota has raised questions regarding the performance and possibly safety of steel truss bridges. The failure investigation indicated that the collapse was due to a design error in the gusset plates used to connect the truss members. The findings also recommend that evaluation of truss connections will be needed, which is a deviation from past practice where only members were considered. There are large numbers of steel truss bridge connections that will need to be evaluated. Future inspections and evaluations of steel truss bridge gusset connections will require effective and efficient methods…

The collapse of the I-35W Bridge in Minneapolis, Minnesota has raised questions regarding the performance and possibly safety of steel truss bridges. The failure investigation indicated that the collapse was due to a design error in the gusset plates used to connect the truss members. The findings also recommend that evaluation of truss connections will be needed, which is a deviation from past practice where only members were considered. There are large numbers of steel truss bridge connections that will need to be evaluated. Future inspections and evaluations of steel truss bridge gusset connections will require effective and efficient methods to collect field data and advanced analysis tools are needed to develop bridge ratings including connections. However, data in the technical literature are lacking on the behavior and performance of representative large-sized gusset plate connections with complex member interactions and multiple connectors. For example, the largest tests reported in the literature are still nearly all less than 1/4 the size of those found in I-35W.  In addition, the number of fasteners in these tests is even less, sometimes only 15% to 20% of those found in large, but not uncommon, truss connections. In addition, actual in-situ failure modes, which may be much more pronounced in larger connections, may not be fully understood or captured in tests of smaller specimens contained in the literature. Current work underway at FHWA’s Turner-Fairbank Highway Research Laboratory will produce new data on large-size gusset plate performance; however, one area that will require additional experimental data is a parametric study of the sway buckling potential of gusset plate connections, as well as the role of corrosion damage, and other initial imperfections.

Bridge inspectors, often the first line of defense in identifying problems, lack the tools to collect geometric and condition data and guidance is needed to aid inspectors in recognizing possibly distressed connections. Rating engineers need validated analytical tools to rapidly and effectively evaluate gusset plate connections thereby enabling rational bridge rating. Current methods rely on specification-based design approaches that are highly simplified and may not be directly applicable to many existing large, complex connections. More sophisticated analytical techniques lack consensus in the profession. Other techniques, such as nonlinear finite element analysis may be very useful in establishing the ultimate strength of these complex connections. However, such methods have not been calibrated with full-scale experimental results, are time consuming to develop, and require specialized training. Finally, modern digital imaging and measurement tools are now available to facilitate data collection and the information that can be produced from these more advanced analytical and experimental techniques would prove to be extremely useful for practitioners, researchers, and inspectors.

Experimental and analytical research is proposed. The objectives of this research are to:

1) Develop advanced tools to rapidly collect field data on gusset plate connection geometry and condition.
2) Develop open-source finite element software and to perform nonlinear analysis and rate connections directly from available drawings or field data.
3) Develop experimental data on sway and non-sway buckling of realistic large-size complex steel truss gusset plate connections.
4) Validate analysis techniques or develop new and improved methods for effective rating and evaluation of gusset plates based on new experimental results and archival data. 
5) Apply experimental meta-data and analytical methods to produce recommendations for inspection and rating of existing gusset plate connections.

Both analytical and experimental techniques will be used to meet these objectives. Current general purpose FE packages are complex and expensive to run and maintain. Most state DOTs do not have the resources to operate these programs. A directed FE analysis software package for conducting nonlinear plate analysis with incorporation of the salient geometric and material properties as well as initial imperfections will be developed. Results will allow automatic meshing, scripted parametric analyses, visualization of results, and reporting features that present summary results in terms of rating factors. The analysis results will be compared with industry standard FE packages and also calibrated with available archival and emerging experimental results. New experimental data will be developed for simulated gusset plates to characterize plate behavior under compression loading to consider the stiffness of the connected member, plate initial imperfections, and thickness and materials on sway and in-plane plate buckling of bridge type gusset plate connections. These data are currently lacking and needed for calibration of the proposed or other analysis methods. Detailed FE models will be used to extend the experimental results considering repair alternatives and effects of corrosion. These new tools proposed will provide a complete package that can collect data from the field or available structural drawings and rapidly produce high-fidelity connection ratings.