| The analysis of pile driving needs to answer one or more of the following questions: (1) What is the static bearing capacity of the pile given observations taken during pile driving? (2) Can the pile be driven, given a complete description of pile, soil and hammer properties (driverability)? (3) Is the pile structurally sound (pile integrity)? (4) What are the stresses in the pile during driving? (5) What is the efficiency of the driving system?; This research work concentrates on the first question: the static capacities of piles, although some of the research outcomes can be applied to the other questions.; A probability based energy approach for effective pile driving control is presented. The energy balance principle is based on a more rational soil-pile interaction model, Smith Model. The Smith model parameters are treated as random variables. The statistical study of Smith model parameters are performed based on the CAPWAP analyses of the 204 pile test results. The statistical values, mean value and standard deviation, and the distribution model of the Smith model parameters were obtained.; A Bayesian updating algorithm for the proposed probabilistic energy method was developed, based on additional observational data in the field, including one or two static load test results or site-specific Smith model parameters.; A new wave equation based approach to determine the static capacities of piles was developed. Its calculation process consists of two schemes: Back-Calculation and Forward-Calculation. The Back-Calculation scheme identifies the soil-pile interaction model parameters from the High Strain Test (HST) data. The Kelvin model is employed to represent the soil-pile interaction. The two unknown model parameters are spring constant K, and the damping coefficient C. The Forward-Calculation scheme uses the identified soil-pile interaction parameters to determine dynamic pile responses to the driving. The Smith damping factors were, then, incorporated herein to determine the static component of the dynamic pile response. The Smith damping factors were viewed as random variables. The forward calculation scheme is cast in a probabilistic framework. |
