In this study, the shaking table tests were conducted to investigate the seismic response of a high-filled reinforced embankment supported by pile and slab structure on slope terrain. The macroscopic damage phenomena of the test model, acceleration response, displacement, dynamic earth pressure and bending moment of the pile were thoroughly examined and discussed. The results revealed that the high-filled subgrade reinforced embankment had a favorable seismic stability. Despite the absence of collapse after 1.2 g seismic load, there was a certain extent reduction in structural resonance frequency. The dynamic earth pressure behind the pile initially increased from the top to the bottom and subsequently decreased near the soil boundary. However, with the seismic magnitude increasing, the peak value of the earth pressure near the pile bottom gradually increased due to pile rotation. The bending moment of the pile presented a bow-shaped distribution. The acceleration exhibited a notable amplification effect along the height of model, while the horizontal acceleration amplification factor decreased with seismic magnitude. Furthermore, the time-frequency domain characteristics and energy distribution of the model were investigated using the Hilbert-Huang Transform. This study provides a theoretical basis for the design of supporting structures for high-filled subgrades in high-intensity earthquake areas.

Considering the engineering background of the dangerous western mountain railroad, large-scale shaking table model experiments were conducted on embankment slopes supported by single and double-row piles, subjected to El-Centro wave excitations. Based on parameters such as displacement and acceleration, an in-depth investigation was conducted to study the differences in dynamic response characteristics between the two slope models. Moreover, the reasons for the differences between the two slopes were explored using fast Fourier transform (FFT) spectra. The results revealed that both the support effect and the differences in anti-slip piles gradually increased with the increase in the input wave amplitude. At input wave amplitudes of 0.1g-0.3g, both single and double-row pile slopes remained stable, with minimal differences in their overall dynamic response characteristics. However, at an input wave amplitude of 0.4g, significant differences in the dynamic responses of both slopes emerged. Macroscopic damage was more apparent in the single-row pile slope, with high slope surface displacement, accumulated soil damage, and noticeable nonlinear characteristics. At an input wave amplitude of 0.5g-0.6g, both slope models exhibited a pronounced elevation effect in the peak ground acceleration (PGA) amplification factor. Additionally, plastic zones were observed on the road cut face and behind the piles in both models. The presence of retaining piles effectively suppressed the upward trend of PGA amplification coefficients along the slope and prevented the connection of plastic zones on the slope surface. Notably, the PGA amplification effect of the single-row pile slope was pronounced, with a wide and deep plastic zone, severe local instability, and relatively weak seismic support effect. The introduction of the FFT spectral ratio revealed that the difference in amplitude amplification effects of single and double-row pile slopes in the 5-10 Hz band was the main reason for the difference in their dynamic responses. Under seismic loading, the failure process of the single-row pile-supported slope involved three stages: initial stability of the slope, plastic deformation of the slope surface soil, and local collapse and disintegration of the slope. In contrast, the double-row pile-supported slope experienced the first two stages of this failure process.