Objective In marine and river environments, local scour around pile foundations can lead to severe damage to the structures, resulting in significant economic losses and potential casualties. Over the past few decades, many researchers have proposed various models to predict scour depth in sandy soils and have systematically evaluated their effectiveness. However, studies on local scour around pile foundations in clay soils are relatively limited. To address this gap, the aim of this article is to compare existing predictive models using experimental data, analyze their rationality and applicability, and provide guidance for engineering construction.

Method This study compiled experimental data from 337 cases of local scour around piles in clay and conducted comparative analyses of 13 different predictive models to explore the impact of clay content, water content, Froude number (F_r) and Reynolds number (R_p) on scour depth. Based on the comparison results, several models with high predictive accuracy and strong applicability were selected.

Result There are significant differences among various prediction models; clay content and Froude number have a substantial impact on the depth of scour, whereas the effect of water content is relatively minor. The applicability of Reynolds number in prediction models is limited due to its dependence on scale effects and flow velocity.

Conclusion This paper enhances the understanding of local scour phenomena around pile foundations in clay soils, and offers guidance for selecting more accurate scour depth prediction models in engineering practices.