Acoustic Emission Precursors in Pile-Reinforced Loess Landslides: A New Early-Warning Signals Identification Approach

Monitoring landslide displacement and anti-slide pile damage is critical for assessing the stability of progressive loess landslides. To address the challenge of capturing precursor information for loess landslide instability under anti-slide pile reinforcement, this study systematically investigates the damage evolution process of slides (through their "slide-stability-reslide" cycles) and anti-slide piles under acoustic emission (AE) monitoring. Cyclic loading tests were employed to simulate the movement of progressive loess landslides. Based on the core causal logic that "slide displacement induces pile damage, damage generates AE signals, and signals invert displacement status", a laboratory-scale physical model was designed to simultaneously monitor slide displacement, pile stress, and AE signals. The research results indicate that the dominant frequency and amplitude of AE signals are significantly correlated with slide displacement: with cyclic loading, both the dominant frequency and amplitude exhibit a "low → high → low" characteristic, corresponding to "low/medium-frequency low-amplitude", "medium/high-frequency medium-high-amplitude" and "low-frequency medium-high-amplitude" signals in the three stages of slide deformation, respectively. The Kaiser and Felicity effects effectively monitor pile damage, and the decrease in Felicity ratio serves as a precursor for landslide early warning. Research results can provide a new methodological framework for early warning systems in pile-reinforced loess landslides.