189 Transportation Research Record: Journal of the Transportation Research Board, No. 2313, Transportation Research Board of the National Academies, Washington, D.C., 2012, pp. 189–197. DOI: 10.3141/2313-20 I.-S. Ahn, 117, and J. O’Connor, 115, Red Jacket Quad, Multidisciplinary Center for Earthquake Engineering Research, University at Buffalo, Buffalo, NY 14261. Current affiliation for I.-S. Ahn: Department of Civil and Environmental Engineer- ing, University of California, 3118 Ghansi Hall, Davis, CA 95616. Y. Park, Depart- ment of Civil Engineering, Bucheon University, Bucheon, Gyeonggi-Do, South Korea 420-735. S. S. Chen, Department of Civil, Structural, and Environmental Engi- neering, University at Buffalo, 226 Ketter Hall, Buffalo, NY 14260. Corresponding author: I.-S. Ahn, ahn.ilsang@gmail.com. was used at the bottom as a tension flange. Four 24-in. (610-mm) deep FRP panels were prefabricated and installed in 1 day on modular abut- ments as part of a demonstration of accelerated bridge construction by the owner, the Department of Public Works, Erie County, New York. The FRP panels were similar to adjacent concrete box beams in that they were joined with shear keys. Each FRP panel tended to behave as a beam in the longitudinal direction. A review of design documents and load rating reports revealed no significant difference in stress and deflection between a simplified analysis and a more detailed finite element analysis (FEA) (4, 5). With respect to live load effects, the design report showed that a grillage analysis was employed to directly consider the load distribution without the use of live load distribution factors (DFs). In the present study, DFs of FRP panels were evaluated from strain values measured in the load tests and calculated from a companion FEA. These DFs were intended to be used in simplified designs of bridges similar to the one presented here. In accordance with Zokaie (6), the DFs in AASHTO Load and Resis- tance Factor Design Bridge Design Specifications (7 ) were developed on the basis of numerical analyses of a suite of bridge models, and regression analyses of them. The maximum girder moment and shear force from numerical analyses were compared with those from sim- plified beam analyses. With respect to FRP composite deck systems, Moses et al. investigated moment DFs of a glass FRP bridge deck system on steel girders (8). It was concluded that DFs of the glass FRP deck system were larger than those for the concrete deck specified in AASHTO Load and Resistance Factor Design Bridge Design Specifi- cations as a consequence of the increased transverse flexibility of the deck. Fu et al. studied DFs of interior stringers of an FRP deck of a truss bridge (9). The DF from FEA was slightly smaller than that from the field test, whereas the DF from AASHTO Specifications was 1.5 times greater than it was from the FEA. Various types of FRP superstructures exist, and each of them behaves quite differently. Consequently, it is a challenging task to develop practical formulae for DFs that cover a wide variety of FRP superstructures, and the present study intended to provide meaningful data for further investigation of this topic. OVERVIEW OF SUBJECT BRIDGE The bridge was built in 2004 in the town of North Collins, Erie County, located in western New York State, and is owned by Erie County. It carries New Oregon Road (a local rural road) over a tributary to the south branch of Eighteen Mile Creek. The traffic volume is approxi- mately 119 vehicles per day, 6% of which consists of trucks. The width of the bridge is 28 ft 10 in. (8.788 m) and its length is 30 ft Live Load Distribution of Hybrid Fiber-Reinforced Polymer Composite Superstructure on the Basis of Field Test Il-Sang Ahn, Younghoon Park, Stuart S. Chen, and Jerome O’Connor Transverse live load distribution in a concrete–fiber-reinforced polymer (FRP) composite hybrid bridge was investigated both experimentally and analytically. The subject bridge was a concrete–FRP composite hybrid bridge, constructed with four 31-ft (9.4-m) long, prefabricated superstructure panels, which were joined along longitudinal field joints made of steel-reinforced grout. The live load distribution factors (DFs) of this bridge were evaluated with strain values measured in field tests and produced from finite element analysis. Design documents and load rating reports showed that a simplified method on the basis of beam- bending could be used in the design process, which required live load DFs. From strain values measured during load tests, the DFs were inferred to be 0.35 and 0.49 for interior and exterior panels, respec- tively. The comparison of load tests done 6 years apart indicated no significant change in the structural performance of the bridge. Although the design of this structure was unique, the evaluation showed that other hybrid structural members also could achieve good load distribution. The investigation indicated that composite structures could be durable with little maintenance required. Fiber-reinforced polymer (FRP) composite material has a high strength-to-weight ratio, electromagnetic neutrality, excellent fatigue behavior, superior durability, and excellent corrosion resis- tance. Since the installation of the first superstructure in 1996, the use of FRP composites for bridges has grown rapidly in the United States (1, 2). The application of prefabricated FRP components on bridges can be divided into two categories: FRP composite deck systems and FRP superstructures. Recently, other materials have been combined with FRP composites to increase the efficiency of the system, which results in a hybrid structure as described in Warn and Aref (3). The superstructure of the simple-span bridge in the present study was made of four concrete–FRP composite hybrid panels (FRP pan- els), in which concrete was used at the top for compression and FRP