Prediction of Pile Running in Layered Soils Based on CPT Data

Abstract

Predicting pile running in layered soils is a critical challenge in geotechnical engineering, especially for large-diameter piles. This study proposes a practical framework for predicting pile running velocity and depth by evaluating pile resistance derived from in situ cone penetration test (CPT) data, including cone tip resistance ( q _c ) and sleeve friction ( q _s ). The framework incorporates three existing cone tip resistance q _c averaging approaches that either struggle to capture soil resistance transitions across layers or tend to overestimate the influence of adjacent layers, thereby introducing inaccuracies in predicting the base response of large-diameter piles. To overcome this limitation, a new cone tip resistance q _c averaging approach, developed from large-deformation numerical analyses, is introduced for assessing pile resistance in this study. The framework was validated using a detailed field-observed pile running event, which provided valuable real-world data. The new averaging approach demonstrated superior agreement with field results, accurately predicting velocity inflections and running depths, compared to the existing methods. This approach offers a reliable tool for assessing pile running in layered soils and can improve design predictions for large-diameter piles.