Bridges that are classified as fracture critical require frequent, intensive inspections resulting in high operation costs. Such costs are justified if the bridge actually poses a risk to the traveling public. However, the assumptions about the classification and inspection frequency of fracture critical bridges are simplified and may be more conservative than necessary. This premise is supported by several historical events where failure of main load-carrying members demonstrated that steel bridges have significant reserve load carrying capability and can be damaged without necessarily resulting in bridge collapse. It makes sense that the classification of a bridge as fracture critical is based on its redundancy, but customary means of establishing redundancy are over simplified, particularly for two-girder systems. An accurate and reliable means of assessing redundancy is needed. It also makes sense to conduct arms-length inspections of fracture critical bridges, but present inspection frequencies, as established by regulation, are arbitrary. In addition to redundancy, the susceptibility of a bridge to fatigue failure is a function of its age, live loading magnitude and frequency, detailing, and materials. Inspection frequencies should be on an engineering consideration of these five factors. Research is needed to improve understanding of these factors and establish an effective means for classifying bridges as fracture critical and for establishing inspection frequencies thereof. Research is needed to characterize and define the redundancies that can be safely incorporated into the evaluation of fracture-critical bridges such that a bridge can be precisely classified and inspection procedures can be optimized. A variety of bridge details must be considered in developing the load paths between the fractured girder the remaining structure. Deck details such as shear stud length, slab haunch, deck reinforcement and top flange thickness as well as cross frames between the girders must be detailed to provide a means of supporting the damaged girder. Findings will not only lead to analysis methods but also to design recommendations to improve redundancy.

To properly classify a bridge as fracture critical and to establish inspection frequencies, the following objectives are targeted: - Achieve understanding of the redundancy of fracture critical systems, such as two-member bridge systems and trusses, including load transfer through the deck to other parts of the bridge and the associated influence of details such as shear stud length, slab haunch, deck reinforcement, top flange thickness, and bridge rail continuity - Based on redundancy findings, establish methods to assess redundancy, including both numerical modeling and simplified methods - Establish the influence of the larger critical crack size that high performance steel can provide on the level of inspection required for fracture critical bridges - Update the relationship of age and loading on the life of fatigue details based on experience and research since the original S-N curves were established The research will include Guidelines for modeling and use of more simplified methods will be developed that can be used to evaluate the behavior of steel bridges with critical structural components.

Researchers will meet with contributing states to understand and prioritize the types of bridges to be studied. Redundancy will be studied with nonlinear structural modeling coupled with laboratory testing to validate analysis predictions. Using probabilistic methods, the research will determine and recommend loads that should be used to check the redundancy limit state. Laboratory and live traffic field tests will be conducted to improve understanding of the load-age relationship on fatigue details. Full size laboratory testing will be conducted to determine the improvement in performance that can be achieved by using high performance steel, especially at cold temperature.

This pooled fund study will be a follow-up / expansion of an existing TxDOT research project being conducted by Drs. Karl Frank and Eric Williamson at the University of Texas, Austin. The pooled fund project will expand from studying redundancy of tub girder systems to include additional bridge types, such as trusses, pier caps, and twin I-girder bridges. The TxDOT project is currently scheduled to continue through fiscal year 2008, at an annual cost of approximately $200,000. If enough states commit funds to expand this research, the TxDOT project will be closed at the end of fiscal year 2007, and approximately $200,000 of TxDOT funds will be added to the pooled fund project, to bring the total project funding to approximately $800,000. Suggested contribution from each state is $25,000 each year, for 4 years ($100,000 total).