Due to concerns with corrosion, the use of fiber reinforced polymer (FRP) as a replacement to conventional steel reinforcement has greatly increased over the last decade. However, elastic modulus values of some commercially available FRP reinforcement hardly reach 20% of that for conventional steel. Existing code relationships for conventional steel have been modified to address this proprietary difference but the modifications are restricted by the empirical nature of the expressions. This paper provides an alternate approach to estimate the deflection of concrete beams by considering effects of tension stiffening that incorporate material properties of the reinforcement as well as the effects of concrete non-linearity in compression. A database containing experimental load-deflection records from 139 glass FRP (GFRP) and 48 carbon FRP (CFRP) reinforced concrete beams was used to calibrate a tension stiffening model for the proposed approach and establish its accuracy as well as precision through statistical analysis. Results were compared to those obtained from existing relationships and indicate that the revised approach provides higher accuracy at service conditions ranging from 25% to 80% of ultimate.

To understand and predict the effect of externally bonded reinforcement (EBR) on the serviceability behavior of FRP (fiber-reinforced polymer) strengthened members, four-point bending tests have been executed on reinforced concrete (RC) beams with span length 3.8 m (150 in.). This experimental campaign was further complemented with tests on strengthened tensile members. These so-called ‘tension stiffening’ tests typically consist of a tensile test on a reinforcing bar embedded in a FRP strengthened concrete prism. As the FRP EBR increases the stiffness of the beams and as a denser crack pattern with smaller crack widths is obtained, the serviceability limit state (SLS) of the strengthened members is positively influenced. Hereby, the behavior in terms of deflection and crack widths can be predicted in a fairly accurate way.

This paper presents detailed investigations into the effective moment of inertia for concrete beams prestressed with aramid fiber reinforced polymer (AFRP) tendons, including an assessment of the existing predictive methods. A three-dimensional nonlinear finite element analysis (FEA) model is developed, based on three different experimental programs reported in literature, to predict the effective moment of inertia of concrete beams prestressed with AFRP tendons. The investigation includes the effect of different sectional properties and various prestressing levels in the tendons. The solved FEA models are compared with several predictive models. The prestressing level in the AFRP tendons significantly influences the transition of the moment of inertia from uncracked section (Ig) to fully-cracked section (Icr) . The existing design standards may not be applicable for beams having a large Ig/Icr ratio (typically over 50) with a low level of prestress (e.g., below 40% ultimate).

Serviceability related to deflections and cracking often controls design of fiber reinforced polymer (FRP) reinforced concrete. The existing approach prescribed in ACI 318 for computing deflection of steel reinforced concrete overestimates member stiffness when FRP is used as the reinforcement. Deflection is then underestimated. Numerous proposals have consequently been made for computing deflection of FRP reinforced concrete, and have mostly involved modifications to Branson’s original ACI expression for the effective moment of inertia Ie. This paper reviews the different procedures used in the past by ACI Committee 440 to compute deflection of FRP reinforced concrete members, and discusses deficiencies of past and present ACI 440.1R deflection calculation guidelines. A case is made for the need to adopt a more rational approach to compute deflection and the basis for proposed changes are reviewed and explained in detail. A statistical comparison of past, present, and proposed approaches for computing deflection are compared with an experimental database that justifies the need for a more rational approach to computing deflection. The paper ends with a clear set of deflection procedures for carrying out serviceability design of FRP reinforced concrete.

An experimental study was conducted to evaluate the increase in crack width occurring over time in FRP-reinforced concrete as a result of sustained loading. Twelve beams (eight GFRP, two CFRP, and two steel-reinforced) were maintained under a constant sustained service load for nearly three years. Three flexural cracks were monitored on each beam over the duration of the test. The observed increase in flexural crack widths over the study was greater in the FRP-reinforced specimens than in the steel-reinforced specimens. On average, flexural crack widths in FRP-reinforced concrete specimens were observed to double over one year of sustained loading. A simple design approach, based on modification to the existing ACI 440 (2006) crack control procedure, is proposed to account for this observed increase in crack widths with time.