Abstract
The current provisions for development length in the ACI 440.11 code disregard the confinement effect provided by stirrups on the bond strength of longitudinal bars and require splice lengths that pose implementation challenges. Given the significant improvement in GFRP material properties, this study investigated the bond strength of sand-coated GFRP bars and proposed a new factor to include the effect of stirrup confinement on the bond-strength provisions. The experimental program involved 16 GFRP-reinforced concrete (RC) beams having a width of 300 mm, and depth 440 mm, consisting of two repetitions for every configuration, subjected to four-point loading. The test parameters comprised lap-splice length and stirrup spacing in the lap-spliced zone. Out of 16 GFRP-RC beams, two beams were reinforced with two M16 (No. 5) continuous bars and six with varying lap-splice lengths [i.e., 40, 60, and 80 bar diameters (d(b))] without confining stirrups. To evaluate the effect of confining stirrups, eight beams were reinforced with two M16 (No. 5) lap-spliced longitudinal bars (i.e., 40 and 60 d(b)) and M13 (No. 4) stirrups spaced at 100 mm (4 in.) and 200 mm (8 in.) center-to-center. Based on experimental results, stirrup confinement clearly increased the bond strength, reduced longitudinal bar slippage, and increased splitting stress. The beams with a splice length of 60 d(b) and stirrups on 100 mm (4 in.) centers achieved 57% higher capacity than those with the same lap-splice length but without stirrups. Further, the ACI 440.11 equation overestimated the bond strength of sand-coated GFRP bars but yielded conservative results with closely spaced stirrups. CSA S6:25 predicted bond-strength values that were close to the experimental results compared to CSA S6:19, and CSA S806:12.