Performance Evaluation of SPT-based Design Methods for the Axial Capacity of Driven Piles in Glacial Deposits-Article

Article

Pdf Summary

The study evaluates seven standard penetration test (SPT)-based methods for predicting the axial capacity of driven steel piles in Ontario’s heterogeneous glacial deposits. The analysis used load-test data from 52 H-piles and pipe piles, along with site soil profiles and SPT measurements. Pile failure loads were identified using the L2 method, and Monte Carlo simulations were used to calibrate ultimate-limit-state resistance factors under LRFD.<br /><br />Most methods showed substantial uncertainty and tended to overpredict pile capacity. Across methods, predictions commonly had high variability, reflecting the difficulty of characterizing glacial soils—particularly their inconsistent gravel content and often very high, erratic SPT values. Five methods notably overestimated capacity, while the combination of the Tomlinson and Berezantzev methods gave roughly unbiased or conservative predictions. The locally developed Jesswein and Liu method, which accounts for soil classification and gravel content, produced the least variable and least biased estimates. Its performance was also supported by separate load-test data from other North American sites.<br /><br />Calibrated resistance factors ranged from 0.05 to 0.66 and were generally lower than AASHTO recommendations, indicating elevated design uncertainty for piles in glacial deposits. The findings emphasize the value of using design methods suited to local ground conditions and verifying designs through pile testing. Results should be interpreted with caution because the study used a limited database and did not account for several potentially important factors, including spatial soil variability, pile plugging, pore pressure, setup time, and seasonal effects.

Keywords

SPT-based methods

driven steel piles

axial capacity

glacial deposits

load-test data

LRFD

resistance factors

Monte Carlo simulation

pile failure loads

soil variability