Keywords: numerical model, piles, load test, ultimate bearing capacity, stress-strain curves, load transfer curves.

It is a common practice to rely on empirical relationships in order to compute the ultimate bearing capacity of deep foundations. In particular, when dealing with piles it is essential to calibrate the load-displacement response, both for axial and lateral loading. Validating theoretical results with field tests enables uncertainties to be reduced and less expensive structures to be designed. The objective of this paper is to simulate the mechanical response obtained from a load test, which include both axial and lateral load using three-dimensional finite element models. Both the axial and lateral load test piles were instrumented to monitor the reaction along the shaft and tip and lateral deflections respectively. The test site was located in the so-called hill zone in Mexico City. This area is comprised by well-cemented tuffs and volcanic materials, interbedded with alluvial sands of medium to high relative density. The soil-pile system was represented with three-dimensional finite element models. Both the soil and the pile were represented with fifteen node axi-symmetric solid elements when modeling axial load, whereas for lateral loading both the soil and the pile were simulated with solid elements and the model was fully three-dimensional. The stress-strain behavior of the soil was described using a bilinear Mohr-Coulomb constitutive law. For cyclic loading, changes in the load direction can lead to permanent deformations in the soil, which were captured by the bilinear Mohr-Coulumb constitutive law used in these models. Piles were considered as an elastic material. The axial model reproduces relatively closely the load-settlement behaviour. The strain gauge located near to the end of the pile indicates a small contribution of the pile tip in the ultimate bearing capacity. Numerical predictions also indicate a little participation of the pile tip. The discrepancy between the results obtained with the finite element model and the load test values for lateral loading can be attributed to the soil cracking observed during the test set up installation. In general good agreement between the measured and computed response was observed.

Comparisons between measured and computed responses indicate that the proposed finite element models can be used to predict the load variation acting over the pile, as well as the ultimate load and maximum deformations for both axial and lateral loading. Thus, this type of formulation can be used safely in static soil-structure interaction analyses. Improvements and advances in ultimate bearing capacity predictions are only possible through calibration with load tests conducted to failure. Obtaining t-z, Q-z and p-y curves for clayey silts, sandy silts and sand with gravels from the numerical model provides useful information to be incorporated into soil-pile interaction analyses.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £145 +P&P)