Evaluation of 18 Direct CPT Methods for Estimating the Ultimate Pile Capacity of Driven Piles

Abstract

Cone and piezocone penetration tests (CPT, PCPT) are widely acknowledged to be useful and powerful in-situ tests for soil classification and characterization, and for evaluating different soil properties, such as strength and consolidation parameters. Due to similarity, between the cone and the pile penetration, CPT data have been used effectively for estimating ultimate pile capacity. Researchers have developed various direct CPT methods to estimate the ultimate capacity of piles (Q_p) from CPT/PCPT data (tip resistance and sleeve friction) with depth. In this study, the measured ultimate pile capacities (Q_p) obtained from static load tests on 80 square precast prestressed concrete piles in Louisiana were used to evaluate 18 direct pile-CPT methods for estimating ultimate pile capacity. Two approaches were used. In the first approach, three criteria (best fit line, arithmetic mean and standard deviation, and cumulative probability of Q_p/Q_m) were adopted, and the sum of ranking of all criteria was used to determine the final ranking of each method. A second approach, multidimensional unfolding, was used to display the ranking data in a two-dimensional space. This approach helps to reveal the typical ranking of the pile-CPT methods, the extent of agreement between the piles, the existence of outliers among the piles, and the similarity between the CPT methods. Based on the results of this study, Bustamante and Gianeselli (LCPC), probabilistic, UF, Philipponnat, CPT2000, UWA, De Ruiter and Beringen, and Schmertmann were found to be the best CPT methods (in order) for estimating the ultimate pile capacity of driven PPC piles.