Validation and Implementation of Bridge Design Specifications for Barge Impact Loading

Since 1991 in the United States, the design of highway bridges to resist collisions by errant waterway vessels has been carried out in accordance with design provisions published by AASHTO. These provisions have remained largely unchanged for more than 20 years, while numerous studies in recent years—conducted by researchers at the University of Florida (UF) and the Florida Department of Transportation (FDOT)—have greatly improved upon the analysis procedures in the AASHTO provisions. The focus of the work discussed in this report was the experimental validation of an improved UF/FDOT barge impact load-prediction model and the implementation of numerous other UF/FDOT procedures into a comprehensive risk assessment methodology that can be readily adopted for use in bridge design. To validate the UF/FDOT barge impact load model, two series of impact experiments were conducted, in which reduced-scale replicas of a typical barge bow were impacted by a high-energy impact pendulum to produce large-scale barge deformations. In support of the validation effort, a material testing program was carried out in order to characterize the strain rate-sensitive properties of steel materials from which the reduced-scale barge specimens were fabricated. Steel specimens were tested in uniaxial tension at strain rates covering seven orders of magnitude. To conduct high-rate material tests, a novel test apparatus was designed and employed that used an impact pendulum to impart the required energy. Data from the material testing program were used to develop constitutive models that were used in finite element barge impact simulations. Additionally in this study, a revised vessel collision risk assessment methodology was developed that incorporates various new UF/FDOT analysis procedures. The complete methodology was demonstrated for two real-world bridge cases, and the results were compared to the existing AASHTO risk assessment method. For these two cases, the revised procedure was found to predict higher levels of risk than the AASHTO procedure. However, it was also noted that the terms in the current AASHTO annual frequency of collapse (AF) expression that are associated with the probability of an impact event occurring may, in fact, overpredict this probability. Consequently, potential inaccuracies in these probability-related terms may account for the higher estimates of risk produced by the revised procedure.