This paper investigates the behavior of circular hollow ultra-high-performance concrete (UHPC) columns reinforced with high-strength steel (HSS) tubes or reinforcements under combined loads, addressing a significant research gap in the application of UHPC for wind turbine towers. Previous studies have primarily focused on conventional towers, lacking comprehensive comparisons of the performance of UHPC columns with various HSS confinement forms such as hollow reinforced concrete (HRC), concrete-filled double-skin steel tubular (CFDST), internally confined hollow reinforced concrete (ICHRC), and externally confined hollow reinforced concrete (ECHRC). To bridge this gap, an experimental study was conducted on four different confinement forms of hollow circular UHPC columns: CFDST, ECHRC, ICHRC, and HRC. These columns were subjected to combined loads of constant compression and cyclic bending. The performance of five specimens was assessed by comparing failure modes, hysteresis curves, ultimate capacity, stiffness degradation, ductility, and energy dissipation, identifying cost-effective confinement forms. The study revealed that UHPC-HSS composite columns exhibit characteristic bending damage modes, with external steel tubes significantly enhancing their ultimate capacity and energy absorption. Replacing the internal HSS tube with reinforcements increased the load-bearing capacity by up to 5.6%. Substituting external steel tubes with steel bars leads to an 18%-27% reduction in load-bearing capacity. In terms of ductility, CFDST columns showed notable improvements, with a 2.8% increase over ECHRC column and a 7.53% increase over ICHRC column. The ECHRC column was identified as the most cost-effective, while CFDST columns, although high-performing, incurred higher costs. Finite element simulations of specimens were developed and validated the experimental results, with errors under 10%. Finally, the feasibility of predicting the ultimate strength of the columns using existing methods was evaluated based on the experimental data.
Read full abstract