Axial Load Capacity Predictions of Drilled Displacement Piles With SPT- and CPT-based Direct Methods-Article

Article

Pdf Summary

The study evaluates methods for predicting the axial capacity of drilled displacement piles (DDPs), which displace rather than remove surrounding soil during installation. The authors analyzed 129 static load tests from 55 construction sites, covering sandy, clayey, and mixed soil profiles, and compared measured or interpreted failure loads with predictions from direct CPT- and SPT-based methods.<br /><br />For tests that did not reach clear failure, Van der Veen’s interpretation method most closely matched observed failure loads. Butler–Hoy and L1–L2 methods also performed relatively well. The original Davisson method and De Beer’s criterion generally underestimated failure, while Decourt and Chin–Kondner tended to overestimate it. A modified Davisson method, using a factor of four for the soil-quake component, improved its accuracy for DDPs.<br /><br />Most capacity-prediction methods underpredicted field performance, likely in part because pile installation improves or densifies surrounding soil—an effect not fully captured by methods based on pre-installation site data. Among CPT methods, Brettmann and NeSmith performed best in sand, Bustamante and Gianeselli in mixed soils, and Eslami and Fellenius in clay. O’Neill and Reese gave the closest SPT-based predictions for sandy and mixed soils, but SPT results showed substantial variability; the authors therefore advise caution in relying on SPT-based estimates.<br /><br />The findings highlight that prediction accuracy depends on soil type, installation effects, and reliable site data. The authors recommend methods suited to DDPs and emphasize that further research is needed to account for the soil changes caused by installation.

Keywords

drilled displacement piles

axial capacity

static load tests

CPT-based methods

SPT-based methods

Van der Veen interpretation

Davisson method

pile installation effects

soil densification

failure load prediction