STATIC PILE RESPONSE FROM FULL-SCALE FIELD TESTS

Get Complete Project Material File(s) Now! »

Winkler Spring model

The Winkler Spring model has been developed to invoke a series of independent non-linear springs and parallel dashpots to represent the soil pressures acting on the pile and radiation (elastic) damping. The initial Winkler model was elastic; this was extended using the p-y approach to account for non-linear soil. This is also referred to as the Beam on Non-linear Winkler Foundation (BNWF), or p-y method. The stiffness, or spring constant, is defined by a p-y curve (Matlock, 1970), which describes the non-linear force-displacement relationship of the spring. The p-y relationships for a given soil can be determined by back-figuring data from lateral lead tests, and difficulties arise in finding a similar soil profile. Since independent springs are used, soil continuity is also neglected. It is uncertain how the p-y curves are affected by pile head fixity and the relative stiffness of the pile and soil (Budhu and Davies, 1988). Also, pile-soil contact is assumed to occur over two dimensions, ignoring the three dimensional and radial components. Characteristic p-y curves, in non-dimensional form, developed by Matlock (1970) for soft clays with free water are shown in Figure 2-2.
The gapping phenomenon was represented in the Winkler Spring model by modelling the foundation system as two series of detachable Winkler springs on either side of the pile (Matlock et al., 1978). The soil adjacent to the pile was modelled with zero tensile strength, hence when the force in the spring reached zero, it detached from the pile. Here it had no influence on the foundation system, as would be the case when gapping occurs during lateral loading. Once the force in the spring was no longer tensile, it re-attached back to the pile.
The limitation of the models discussed above is that they primarily model pushover (static) response and do not consider dynamic behaviour. Damping effects have been included in Winkler models by attaching dashpots to the pile, with a similar layout to spring elements. This model type neglects the kinematic response, considering only inertial interaction. Kinematic response was considered by adopting the Beam on Dynamic Winkler Foundation (BDWF). The ends of the spring and dashpot elements are connected to a representation of the free field soil, as opposed to being fixed, see Figure 2-3. The motion of the free field soil serves as the input excitation for the pile-soil system used in the previous models (Matlock and Foo, 1978; Makris and Gazetas, 1992; Pecker and Pender, 2000; Wotherspoon, 2009).

Elastic Continuum Model (ECM)

Poulos (1971a, 1971b) used the Elastic Continuum Model (ECM) to investigate the response of an elastically loaded pile in an elastic soil. Several limitations were encountered in this study. A homogeneous isotropic semi-infinite soil is unrealistic as the soil modulus in reality is likely to vary with depth. The soil is also assumed to behave elastically, this approximation is only valid at very small strains; at large deflections soil behaviour is highly non-linear, thus a linear approximation is not valid. To overcome this limitation, Poulos considered local yielding of the soil to account for the non-linearity of the soil. Randolph (1981) simplified expressions developed by Poulos, which included pile length, by introducing the concept of active length for long, flexible piles. This is the upper length of the pile where significant deformation occurs.
Expressions were valid provided the length of the pile was greater than the active length. When kinematic interaction is considered, vertical shear waves travelling through the soil interact with and deform the pile resulting in significant deformation along its entire length (Fan et al., 1991; Makris and Gazetas, 1992; Kavvadas and Gazetas, 1993). This suggests a finite model is required as opposed to the infinite active length concept. Also, the ECM accounts for soil continuity, addressing the major limitation associated with the Winkler Spring model.
Equations were later developed for the ECM for constant Young’s modulus with depth (Davies and Budhu, 1986), linear increase in Young’s modulus (Budhu and Davies, 1988) and parabolic variation of Young’s modulus (Gazetas, 1991). The displacements and rotations of the pile head are calculated assuming elastic behaviour of the pile and soil, based on the applied horizontal load and moment at the pile head, and flexibility coefficients.
Davies and Budhu (1986) extended the elastic equations to account for the non-linear behaviour of the soil and pile. This introduced yield influence factors to the elastic displacements and rotations. Pender et al. (2012b) developed these equations for cyclic lateral loading of a homogeneous soil profile, and confirmed results with field testing of driven piles in Auckland residual clay and Finite Element software OpenSeesPL (Lu et al., 2010). Wolf (1985) developed an equivalent single degree of freedom (SDOF) model for the dynamic response.

READ Coal Genesis in the Ecca Group

CHAPTER 1 INTRODUCTION
1.1 Overview
1.2 Objectives and scope of study
1.3 Thesis outline
CHAPTER 2 LITERATURE REVIEW
2.1 Overview
2.2 Analytical studies of pile-soil interaction
2.3 Experimental studies of pile-soil interaction
2.4 Code approaches and Soil-Structure Interaction (SSI)
2.5 Summary from previous research
CHAPTER 3 TEST SET-UP AND METHODOLOGY
3.1 Overview
3.2 Pile details and layout
3.3 Pile head preparation
3.4 Geotechnical site conditions
3.5 Instrumentation and data collection
3.6 Data preparation
3.7 Test procedure and equipment
3.8 Test program
CHAPTER 4 STATIC PILE RESPONSE FROM FULL-SCALE FIELD TESTS
4.1 Overview
4.2 Pull-back force-displacement response
4.3 Pile-soil gap measurements and observations
4.4 Summary
CHAPTER 5 DYNAMIC PILE RESPONSE FROM FULL-SCALE FIELD TESTS
5.1 Overview
5.2 Response in the frequency domain
5.3 Response in the time domain
5.4 Hysteretic pile response
5.5 Summary
CHAPTER 6 NUMERICAL MODELLING OF LATERAL PILE RESPONSE
CHAPTER 7 CONCLUSIONS
APPENDIX A PILE 3 DYNAMIC DATA FILE INPUT INTO RUAUMOKO
APPENDIX B EXAMPLE DATA ANALYSIS FILE ON MATLAB