Abstract
Dynamic pile load testing (i.e., PDA and CAPWAP) was performed during the driving of six piles at two sites included large and low displacement H-piles, and actual pile capacity was determined during the end of drive. The load test provided an opportunity to compare pile design techniques to measured pile performance. The soils at one of presented sites prevent the pile driving process from being completed and the required pile length and capacity were not achieved due to early refusal. Therefore, the engineers redesigned the deep foundation system, whereby the large displacement prestressed concrete piles (PCP's) were replaced with low-displacement steel H-piles. In this paper, seven dynamic methods for predicting axial pile capacity of driven piles are investigated and summarized. The dynamic formulas included Eytelwein, Modified ENR, Janbu, Danish, Navy-Mckay, Gate, and PCUBC. The measured pile capacities were compared to the predictive capacities to evaluate which predictive method would be best suited for estimating the pile capacity at site where such difficult soils may encountered. The evaluation revealed that the pile dynamic formulas are mostly under-predicting pile capacity. Amongst the seven methods, the Danish method gave the most realistic values of the pile capacity. The predictions using the Gates and Modified ENR methods were found to be overly lower than the measured values and was ranked least desirable amongst the methods. The predictions at site where early refusal was encountered, found to be overly lower than the measured values. However, concrete piles were replaced by H-pile.