This paper reported a series of hysteretic torsion experiment to investigate the torsional behavior of rectangular hollow reinforced concrete (RHRC) column strengthened by fiber reinforced polymer (FRP). Six RHRC column specimens with different number of longitudinal reinforcements, spacing of stirrup and strengthening method using FRP were designed. One was not strengthened, four were strengthened with CFRP, one was strengthened with CFRP and GFRP. The experimental results showed that the primary failure modes of specimens were the spalling of surface concrete with the detachment of FRP. In details, under the hysteretic torsional load, the interaction between adhesive and concrete caused the intersecting diagonal cracks in the internal concrete. Compared with the hysteretic curve of specimen without FRP strengthening, FRP strengthening can significantly improve the initial stiffness by 50 % and peak torsional strength by 70 %. For RHRC column without strengthening, the fullness was poor because of the weak torsional energy dissipation. The FRP strengthening can also enhance the torsional energy dissipation and seismic behavior of RHRC column. To predict the complex torsional behavior of RHRC column strengthened by FRP, a finite element (FE) model and a constitutive model were developed. The FE model considered potential cracks in concrete and FRP-concrete interface based on the application of the cohesive zone model (CZM), whereas the constitutive model accounted for interface damage and plasticity. The results of the performed simulations indicated that the proposed model can effectively represent the hysteretic mechanical behavior of columns under torsional load, which cannot be achieved using conventional FE methods.
Local scour has been reported as a common phenomenon for pile-group foundations in marine, coastal, and riverine sites, while its effects on the seismic behavior of pile-group foundations are yet to be well documented. This paper reported a pair of quasi-static cyclic loading tests on 2 x 3 scoured pile-group foundation specimens, one in global scour and the other in local scour, to understand the influence of local scour on the seismic behavior, particularly in terms of soil-pile interaction features and failure mechanisms. Test results indicate a flexural failure mode for both specimens. Local scour does not change the order of limit states but postpones the occurrence of concrete cover cracking and rebar yielding due to the reduced lateral stiffness of the locally scoured piles. Moreover, local scour rarely changes the aboveground damage regions at the top of outer piles (one time the side length of squared pile section, D ), but significantly aggravates and deepens the underground damage regions at outer piles (from 4 5D D for global scour to 3.7 6D D for local scour). Besides, local scour reduces the displacement ductility factor from 2.50 to 1.25 for the Easy-to-Repair limit state, resulting in adverse impacts on pile-group foundations in general.