Effective Post-Tensioned Splicing System for Prestressed Concrete Piles

Precast prestressed concrete piles are widely used in deep foundation construction. Due to unexpected site soil conditions and difficulties associated with transportation and handling long precast piles, splicing is sometimes necessary at the construction site. Available splicing methods utilize steel type connections that are more suitable for reinforced concrete construction and result in limited tensile capacity at the splice. This dissertation describes studies associated with the development of a new post-tensioned splicing system using staggered, embedded anchorages. The new system has the potential to provide the same tensile capacity as a one piece prestressed pile. To develop the post-tensioned splicing system it was necessary to conduct varied numerical analyses to solve immediate technical problems associated with the design, fabrication and testing of a prototype. This included the design of a self-stressing prestressing bed, optimization of the shape of the embedded anchorages and their layout within the piles being spliced. The focus of the dissertation is on non-linear finite element studies conducted to model the flexure behavior of prototype laboratory and full-sized spliced piles in comparison to their identical non-spliced counterpart. Though finite element analysis of prestressed elements is not new, issues relating to modeling post-tensioned, spliced elements with embedded, staggered anchorages have not been the subject of any previous investigation and constitute the principal contribution of this study. Nonlinear finite element analysis was conducted using ANSYS. The William-Warnke failure criterion used to establish concrete failure. A three-dimensional analysis was conducted in which SOLID65 element was used for modeling concrete and LINK8 for the prestressing strands. The post-tensioning ducts were modeled using PIPE20 elements. Perfect bond was assumed between the concrete and the ducts. Embedded anchorages were modeled as fixed locations within the concrete. Epoxy used to join the two splicing surfaces was modeled using contact elements. Since the layout of the post-tensioning ducts was staggered, a full model was required. In contrast, advantage was taken of symmetry for the analysis of the one piece controls. The finite element model was able to accurately capture the flexural behavior of both the control and the spliced piles. The results suggested that tensile separation at the splice interface acted as a pivot about which the section rotated. As a result, the compression failure zone in the spliced pile was confined to a smaller region compared to the control. The stress distribution in the spliced pile indicated that the concrete in the cover above the splice was crushed at the ultimate stage before the steel had yielded. As a result, the ultimate capacity of the spliced pile was controlled by concrete failure. The results also indicated that, among the multiple layers of post-tensioning strands, only one approached yield while others remained in the elastic range. As a result, when the applied load was released, the spliced pile rebounded back to a large degree, which resulted in a much smaller residual permanent deformation. This behavior of a spliced pile can be beneficial for structures in a seismic zone because it will induce smaller secondary moments. This study helped to refine and improve the new post-tensioned splicing system. Its availability makes it possible to extend and further improve the concept without the need for costly prototype fabrication and testing