Title: Deflection Behavior of Beams Prestressed with Bonded FRP Tendons

Author(s): Wassim Nasreddine, Peter H. Bischoff, and Hani Nassif

Publication: Symposium Paper

Volume: 360

Issue:

Appears on pages(s): 491-510

Keywords: cracking, deflection, effective moment of inertia, fiber-reinforced polymer (FRP) tendons, prestressed concrete

DOI: 10.14359/51740645

Date: 3/1/2024

Abstract:

The use of FRP tendons has become an attractive alternative to steel tendons in prestressed concrete structures to avoid strength and serviceability problems related to corrosion of steel. There is however a lack of knowledge in serviceability behavior related to deflection after cracking for beams prestressed with FRP tendons. Conventional approaches used to compute deflection of cracked members prestressed with steel is problematic at best, and the situation is exacerbated further with the use of FRP tendons having a lower modulus of elasticity than steel. Deflection of FRP reinforced (nonprestressed) concrete flexural members computed with Branson’s effective moment of inertia 𝐼􀀁 requires a correction factor (called a softening factor) that reduces the member stiffness sufficiently to provide reasonable estimates of post-cracking deflection. For FRP prestressed concrete however, this approach does not always work as expected and deflection can be either underestimated or overestimated significantly.

This study investigates the accuracy of different models proposed for estimating deflection of cracked FRP prestressed members using a database of 38 beams collected from the literature. All beams are fully prestressed. Results indicate that using Branson’s effective moment of inertia 𝐼􀀁 with a generic softening factor can produce reasonable estimates of deflection provided the 𝐼􀀁 response is shifted up to the decompression moment or adjusted with an effective prestress moment defined by an effective eccentricity of the prestress force. The former approach overpredicts deflection by 20% on average while the latter overpredicts deflection by not more than 5% based on the beams available for comparison. Assuming a bilinear moment deflection response overpredicts deflection by 12%, while an approach proposed by Bischoff (which also shifts the 𝐼􀀁 response upwards) overpredicts deflection by 23%. These last two approaches work reasonably well without the need for a correction factor.