Abstract
The development of pultruded glass fiber reinforced polymer (GFRP) bars for internal reinforcement of concrete, together with dedicated limit-state design guidelines, has led to a recent breakthrough in the field of tunnel excavation. The use of GFRP bars in softeyes, which are openings of retaining walls to be penetrated by tunnel boring machines (TBMs) during excavation, is becoming mainstream. The low shear strength and brittleness compared to steel bars facilitate and expedite excavation, resulting in time and cost saving, as well as improved safety. Large-size (#10) GFRP bars are typically used as flexural reinforcement for the massive softeyes, often in bundles. However, the flexural and shear design algorithms adopted by the American Concrete Institute (ACI) for fiber reinforced polymer (FRP) reinforced concrete (RC) have never been experimentally validated with full-scale tests. Question marks exist on potential detrimental effects on the concrete shear strength contribution that accrue from size effect, and on the flexural strength of RC members due to shear lag in the large-size longitudinal reinforcement, and due to the use of bar bundles. In this paper, the fundamentals of flexural and shear design of FRP RC are first outlined. Then, an experimental program that included bending tests on five full-scale softeye beam specimens is presented and discussed. The test matrix was designed to study the shear and flexural response of large-scale members using different layouts of flexural and shear reinforcement. The results demonstrate the validity of the current ACI design algorithms, and back the identification of areas of research to improve their efficiency.