Distribution of live load shears in FRP composite tub girder highway bridges
Published: 2024
Publication Name: Engineering Structures
Publication URL: https://doi.org/10.1016/j.engstruct.2024.119188
Abstract:
In the design of slab-on-girder highway bridges consisting of conventional materials like concrete and steel in the United States, the vehicular live load carried by a single girder is calculated using distribution factors (DFs) defined in the American Association of State Highway and Transportation Officials (AASHTO) design specifications. However, shear DFs for the recently developed fiber reinforced polymer composite tub (CT) girder do not exist within current design codes, and to-date in-service CT girder bridges have been designed using AASHTO shear DFs for concrete box girders. To assess shear live load distribution in CT girder bridges, diagnostic live load tests were performed on two in-service highway bridges under heavy truck loads. High-fidelity finite element (FE) models calibrated to the test results were simplified to reflect conventional design assumptions. The high-fidelity FE models indicated that AASHTO over-predicted live load shears in the most heavily loaded interior girder by as much as 35 %, but can under-predict exterior girder live load shear. Parametric studies using the simplified FE models indicated that while the most influential parameter on CT girder shear DFs is girder spacing, girder bottom flange width can also play a significant role. The simulations and diagnostic live load tests both indicate that the AASHTO shear DF expressions for concrete box, slab-on-girder bridges that are currently used in CT girder design typically over-predict shear DFs for interior CT girders. Simulations with the simplified model indicate over-predictions of DFs for interior CT girders of up to 30 % for longer spans and large girder spacing. However, in the CT girder that experienced the greatest shear strain during field load testing, measured strains in the most heavily loaded web were 22 % higher than the average girder web shear strain, a factor not currently accounted for by existing AASHTO DFs or in CT girder design.