Pile-supported embankments are one of the most commonly used techniques for ground improvement in soft soil areas. Existing studies have mainly focused on embankments supported by end-bearing piles under static loading, with limited research on floating pile-supported embankments under cyclic traffic loading. In this study, model tests for unreinforced floating, unreinforced end-bearing, geosynthetic reinforced floating, and geosynthetic reinforced end-bearing pile-supported embankments were conducted. Cyclic traffic loading was simulated using a three-stage semi-sinusoidal cyclic loading. Comparative analyses and discussions are performed under floating and end-bearing conditions to investigate the influence of floating piles on the soil arching evolution and membrane effect under cyclic loading. The results indicate that floating piles result in earlier stabilization of surface settlement. There is less arching and membrane effect induced by floating piles, and the arching does not continue to degrade under cyclic loading. Less membrane effect in floating pile-supported embankments results in less geosynthetic and pile strain. The degree of membrane effect in floating pile-supported embankment largely depends on the pile-end condition.

This case study aims to evaluate the impact of deep excavation on the adjacent short floating pile and lateral deformation control strategies using capsule expansion technology (CET). Two control strategies, i.e., real-time control (RTC) and one-time control (OTC), were applied to control the lateral displacement of piles. In this case, the wall lateral deflections (delta hm) range between delta hm=0.075%He and delta hm=0.11%He, which are relatively small and less than the specified protection levels. Although the wall deflection was controlled to a relatively small level through reasonable excavation and support schemes, the maximum horizontal displacement of the short floating pile reached 13.2 mm (0.054%He). Therefore, reasonable deformation control measures are necessary. After three stages of RTC treatment, the maximum lateral displacement of P2 was reduced by 49.2%, while P1 was decreased by 22.7% treated by OTC. Meanwhile, multiple RTCs can always control the pile deformation within the cracking limit, which avoids the dilemma of protecting the pile after it has been damaged. It confirms the feasibility and efficiency of CET in controlling pile deformation in real-time. In addition, RTC for pile lateral displacement mainly includes two aspects: (1) expansion directly induces lateral displacement of piles; and (2) expansion compensates for the soil stress loss in front of the pile to reduce the impact of the next excavation on the pile. Therefore, as external influence sources have long-term adverse effects on adjacent piles, RTC as an efficient control method should be given priority consideration for controlling pile lateral displacement.

This study presents two large-scale model tests to investigate the load transfer mechanism of floating pile-supported embankments subjected to cyclic loading. The soft soil and piles were prepared using Kaolin clay and reinforced concrete. Results on cumulative settlements, pile efficacy, and strain distribution were obtained and analyzed under semi-sinusoidal cyclic loading. The results show that the floating pile increased surface settlement by 7.1% compared to the end-bearing pile-supported embankment. The soil arching in floating pile-supported embankment does not degrade under cyclic loading but slowly enhances with settlement development. Floating piles result in less arching, membrane effect, and pile strain.