Bond Performance of GFRP Reinforcing Bars under Uniform or Gradient Temperature Distributions

Standard pullout tests (e.g., 150 mm cubes) with a short embedment length have shown a significant reduction (80% loss at 200°C) in the bond strength of glass fiber-reinforced polymer (GFRP) bars at high temperatures but little information is known about how this information can be extended to longer embedment lengths. To address this knowledge gap, a more detailed bond behavior of GFRP reinforcing bars has been examined in this study with a modified pullout test method in which pullout specimens consisting of GFRP bars embedded in 350 mm (22db) long concrete cylinders were tested at high temperatures. The tests examined the bond degradation under different temperature distributions (i.e., uniform and gradient). The results were compared with the previous pullout tests with 68 mm (4db) embedment length and showed that the pullout capacity of GFRP bars in this test was 3.3, 4.0, and 2.0 times higher at 115°C, 135°C, and 300°C, respectively. The results were the first to prove the effectiveness of longer embedment length particularly at temperatures below 200°C. Additionally, pullout tests with temperature gradients simulated the bond behavior of GFRP bars at the unexposed end zones of GFRP reinforced concrete beams or slabs. The results showed that the tensile force in GFRP bars gradually transferred to the lower temperature zones as the heated parts of the bar lost bond with concrete. The results showed that the concentrated forces over a short length of GFRP bars can trigger concrete splitting and loss of bond even if the GFRP bars still maintain the bond with concrete. Providing thicker concrete cover at the end zones of GFRP reinforced members is recommended to enhance fire resistance. The findings of this study are being used in developing forthcoming FRP reinforced concrete design codes.