Assessment of axially-loaded pile dynamic design methods and review of INDOT axially-loaded design procedure

Abstract

The general aim of the present research is to identify areas of improvement and propose changes in the current methodologies followed by INDOT for design of axially loaded piles, with special focus on the dynamic analysis of pile driving. Interviews with INDOT geotechnical engineers and private geotechnical consultants frequently involved in INDOT’s deep foundation projects provided information on the methods and software currently employed. It was found that geotechnical engineers rely on static unit soil resistance equations that were developed over twenty years ago and that have a relatively large degree of empiricism. Updated and improved static design equations recently proposed in the literature have not yet been implemented in practice. Pile design relies predominantly on SPT data; cone penetration testing is performed only occasionally. Dynamic analysis of pile driving in standard practice is performed using Smith-type soil reaction models. A comprehensive review of existing soil reaction models for 1-dimensional dynamic pile analysis is presented. This review allowed an assessment of the validity of existing models and identification of their limitations. New shaft and base reaction models are developed that overcome shortcomings of existing models and that are consistent with the physics and mechanics of pile driving. The proposed shaft reaction model consists of a soil disk representing the near field soil surrounding the pile shaft, a plastic slider-viscous dashpot system representing the thin shear band forming at the soil-pile interface located at the inner boundary of the soil disk, and far field- consistent boundaries placed at the outer boundary of the soil disk. The soil in the disk is assumed to follow a hyperbolic stress-strain law. The base reaction model consists of a nonlinear spring and a radiation dashpot connected in parallel. The nonlinear spring is formulated in a way that reproduces realistically the base load-settlement response under static conditions. The initial spring stiffness and the radiation dashpot take into account the effect of the high base embedment. Both shaft and base reaction models capture effectively soil nonlinearity, hysteretic damping, viscous damping, and radiation damping. The input parameters of the models consist of standard geotechnical parameters, thus reducing the level of empiricism in calculations to a minimum. Data collected during the driving of full-scale piles in the field and model piles in the laboratory are used for validating the proposed models.