This study investigates the application of machine learning (ML) algorithms for seismic damage classification of bridges supported by helical pile foundations in cohesive soils. While ML techniques have shown strong potential in seismic risk modeling, most prior research has focused on regression tasks or damage classification of overall bridge systems. The unique seismic behavior of foundation elements, particularly helical piles, remains unexplored. In this study, numerical data derived from finite element simulations are used to classify damage states for three key metrics: piers' drift, piles' ductility factor, and piles' settlement ratio. Several ML algorithms, including CatBoost, LightGBM, Random Forest, and traditional classifiers, are evaluated under original, oversampled, and undersampled datasets. Results show that CatBoost and LightGBM outperform other methods in accuracy and robustness, particularly under imbalanced data conditions. Oversampling improves classification for specific targets but introduces overfitting risks in others, while undersampling generally degrades model performance. This work addresses a significant gap in bridge risk assessment by combining advanced ML methods with a specialized foundation type, contributing to improved post-earthquake damage evaluation and infrastructure resilience.

This study provides a comprehensive analysis of the undrained failure envelope for spudcan foundations in anisotropic clays using the AUS failure criterion as the soil strength model. The influence of embedment depth (L/D) and anisotropic strength (re) on spudcan behaviour under combined loading conditions is investigated. Failure envelopes are derived through three-dimensional finite element limit analysis (3D FELA) in both (H/ suTCA, M/suTCAD) and (V/Vult, H/suTCA, M/suTCAD) spaces. The study also illustrates spudcan foundation failure mechanisms, providing valuable insights for designing footings in anisotropic clays under combined loads (V, H, M). Additionally, an innovative soft-computing approach is introduced: a machine learning model that integrates categorical boosting (CatBoost) with the flower pollination algorithm (FPA) for optimized predictions of the spudcan failure envelope. The proposed FPA-CatBoost model is validated against numerical FELA results, demonstrating a strong correlation and offering engineers a reliable tool for determining spudcan foundation failure envelopes under varied loading conditions.