Capacity of driven piles in cohesionless soils including residual stresses

Static formulas based on limit equilibrium theories often provide misleading predictions of pile capacity in cohesionless soils. Researchers have been working within the last few years to improve empirical prediction methods. However, even the most recent methods neglect one or more of four factors; namely: residual stresses due to pile driving, actual soil parameters prior to failure, mean normal stress levels, and stress history. Available indirect methods used for residual stress measurement are not accurate. Furthermore, no accurate method for their prediction, except the wave equation, is available. Therefore, in this dissertation a new method that takes into account all related parameters for residual stress prediction was developed. It was shown that the existence of these stresses is the principal reason for observing limit values of pile tip and shaft capacities. Due to high stress levels beneath the pile tip prior to failure, the initial angle of shearing resistance may be substantially reduced. This may result in serious overestimation of the pile tip capacity, especially in dense sands, if the initial angle were used. In this research, a procedure was developed to account for this behavior. For many of the available methods, the vertical overburden pressure was used for predicting pile capacity. However, this capacity is actually affected by the mean normal stress. Correlations in terms of this stress are proven to be superior to those produced when the overburden pressure was used. Stress history was found to have a major effect on pile capacity in both reducing soil compressibility and increasing lateral pressure on the pile shaft, and hence increasing the pile capacity. These effects have not been taken into account in available prediction methods. In this research, available load test data were properly adjusted for residual stresses, and were then used to develop a new empirical prediction method for driven piles in cohesionless soils. The new method takes into account parameters at failure, mean normal stress, and stress history. This method was indicated to be superior to other available methods by comparison with actual measurements