Improving flexural performance of post-tensioned pre-cast pre-stressed RC segmental T-beams

Abstract

For a segmental pre-stressed beam, the joints between segments will open at the tension side under the effect of bending moment. Accordingly, the concrete will be crushed at the compression side and/or additional tensile force will be developed at the pre-stressed tendons. In addition, the open joint provides easy passage way for environment attack on pre-stressing tendons leading to durability problems. Thus, the overall performance of the segmental beam can be adversely affected. In the current paper, in order to tackle these issues, strengthening techniques were proposed to restrain the joints between segments at tension and/or upgrading the compressive strength at compression side. Thus, five pre-cast concrete segmental T-beams having dry joints pre-stressed by un-bonded tendons in addition to two monolithic control beams (one of them was pre-stressed) were prepared and tested under the effect of an incremental four-point loading scheme. The compression side was upgraded by the replacement of the upper top half of flange thickness by cast-in-situ high-performance concrete (HPC) layer. While the tension side was strengthened by either providing a steel assembly to restrain the open joints by connecting reinforcing steel bars across the joints or splicing the extended longitudinal bars between the segments and then encapsulating them by strain-hardening cementitious composite (SHCC) material. Test results showed that the ultimate capacity of the pre-stressed segmental beams can reach up to 96% of the ultimate capacity of monolithic pre-stressed beam when the segmental beams were strengthened at both tension and compression sides. However, this comes on the expenses of the developed ductility. Using the proposed steel assembly enabled the strengthened segmental beam to develop more ductile response than that obtained when spliced encapsulated bars were implemented. Finally, the recommended un-bonded post-tensioned tendon stress at ultimate stage stipulated by AASHTO, ECP 203-2018 and ACI 318-14 codes were compared against the experimental results. The comparison showed that all considered codes predicted well the un-bonded tendon stress; however, AASHTO equations yielded the least conservative predictions.