Seismic performance and damage identification of anti-slide piles under varying initial damage conditions using wavelet packet energy spectrum

Anti-slide piles are commonly used to stabilise high and steep slopes in earthquake-prone areas in southwestern China. Herein, we investigate the impact of initial damage on the seismic performance of anti-slide piles. For this purpose, we selected a representative slope adjacent to the Jiuzhaigou Bridge in the Sichuan–Qinghai Railway; we employed a three-dimensional dynamic finite element method combined with the local stiffness reduction approach to simulate three different initial-damage scenarios: intact, slightly damaged and heavily damaged. The dynamic displacement, bending moment and shear stress responses of the piles were comprehensively analysed. Using wavelet packet energy spectrum (WPES) analysis, we introduced two indices: the damage index (D_PERV) and its increment (|△D_PERV|). The results showed that both the initial damage and seismic energy control the peak dynamic response of the piles. Specifically, high initial damage accelerates stiffness degradation, leading to large horizontal displacements, whereas intact piles sustain high bending moments and shear forces. The distribution of |△D_PERV| along a pile reveals three post-earthquake performance stages (i.e. minor, moderate and severe), which agree well with the observed mechanical response characteristics and form the basis for targeted reinforcement strategies. The main innovation of this study is the combined use of initial-damage simulation with WPES analysis, thereby establishing a quantitative diagnostic framework (D_PERV and |△D_PERV|) for anti-slide piles. This framework determines the non-linear relationship between seismic response and damage evolution and provides a rapid, usable tool for health monitoring and post-earthquake decision-making in landslide-prone mountainous railway areas.