Characterization of 't-z' model design parameters for augered cast-in-place piles using field load test data.

摘要

With the increasing use of augered cast-in-place piles in new construction, it is important that proper design parameters be incorporated when evaluating pile capacity and performance using reliability-based design methods. This paper focuses on developing "t-z" model parameters from analysis of static axial compression and tension load test data from a project site along the Missouri River floodplain in northwest Missouri. Data was collected from a total of twelve axial load tests (six compression and six tension) and includes dial gauge readings from the pile heads as well as vibrating wire strain gauge data from multiple locations throughout several of the test piles. The "t-z" method has been used extensively as a soil-structure interaction model to evaluate the settlement of deep foundations. The soil-structure interaction modeled in this analysis was based on hyperbolic load displacement behavior using effective (drained) stresses. The development of the "t-z" model parameters has been accomplished using finite difference methods to analyze the non-linear soil-structure interaction along the sides of the piles. During the analysis, the mean shear modulus of soil-structure interface subgrade reaction, Kinit, and the mean ultimate shear strength of the soil-structure interface, tauu, were back-calculated from each set of load test data and were based on the assumption of a single-layer, homogenous soil profile. These "t-z" model parameters were then compared to standard field investigation data, including standard penetration tests (SPT) and cone penetrometer test (CPT) soundings, and effective overburden stress to develop correlations suitable for service limit state design of augered cast-in-place piles. While there was some indication of a linear relationship between Kinit and the field investigation data, there was not a sufficient quantity of data in the analysis to properly identify any statistical trends. The relationship between tau u and the field investigation data was much more variable and did not provide any distinct correlation. The plot of the data relating the model parameters to the effective overburden stress exhibited some grouping but the sample size and distribution was not sufficient to identify any statistical trends.