Fiber-Reinforced Polymer (FRP) Bridge decks offer great advantages in highway bridge rehabilitation and new construction, due to reduced weight and maintenance costs, and enhanced durability and service-life. In practice, however, lack of bridge engineering design standards and guidelines have prevented wider acceptance and application of FRP bridge decks by transportation officials. This dissertation focuses on the study of an engineered FRP deck-steel stringer bridge system through experimental testing and both Finite Element analyses and analytical methods.

A prototype mechanical shear connection was developed and designed to be used with any type of FRP panels that can accommodate any panel heights. This non-grouted sleeve-type connector can secure the deck onto a welded stud and can sustain shear forces at FRP panel-steel stringer interface. Static and fatigue tests were conducted on push-out connection specimens, and later on a scaled bridge model. The strength, stiffness, and fatigue performance characteristics of the connection were fully investigated. Constructability issues were also evaluated, such as ease of installation and economic manufacturing of the connector. Design formulations were established based on the test results.

Following the connection study, a 1:3 scaled bridge model of a honeycomb FRP deck on steel stringers was evaluated. The deck was attached to three supporting steel stringers using the proposed sleeve-type mechanical connections. The model was designed as partially composite to satisfy AASHTO limits and requirements. Several issues were evaluated that included: (1) deck attachment procedures; (2) transverse load distribution factors; (3) local deck deflections; and (4) system fatigue behavior. After the bridge model was tested in the linear range, a 1.2-m wide T-section, of an FRP deck section attached to the middle stringer, was cutout from the bridge model and tested in bending for service and failure loads. The evaluations included: (1) Degree of composite action, (2) Effective deck-width, and (3) service-limit and ultimate-limit states under flexure loads. The behavior of the FRP deck under partial composite action was defined fully by these tests.

Finite element models of the scaled bridge model and T-beam section were formulated using ABAQUS. Besides the experimental tests and FE analyses, analytical solutions were developed to verify the test results. An explicit series solution for stiffened orthotropic plates was used to evaluate the bridge response and obtain load distribution factors of FRP deck-on-steel-stringer bridges. Also a harmonic analysis that was developed for FRP thin-walled sections was formulated to define effective-width for FRP decks as an explicit solution. The outcome of this study was to propose design guidelines and recommendations for FRP honeycomb bridge decks for applications in bridge engineering practice.