Abstract:
This paper reports results of a study to predict load-settlement behaviour of strip foundations on a dry uncemented sand
using only soil stiffness data, under both working loads and extreme loads. This involved predicting the strength of a
foundation system without using any strength parameters. A non-linear stepwise load-settlement prediction algorithm was
utilised, which used the small-strain shear modulus of layers below the footing, a suitable stiffness modulus reduction
curve and elasticity theory. The algorithm was adapted to predict the ultimate bearing capacity, as well as to construct a
stress-settlement curve and determine the bearing pressure experienced beneath a foundation at any particular settlement.
The method was evaluated using two different sets of modulus reduction curves, for small-scale model foundations on
loose, medium dense and dense sand. Model tests were conducted at 1g. Particle Image Velocimetry (PIV) was used to
observe strains within the soil during loading. These strains were used to calculate the shear modulus of discrete layers
throughout loading and back-estimate the layer small-strain shear moduli. Additionally, the mode of shear failure for each
test at ultimate bearing capacity was observed using PIV. Modulus reduction curves that initially degrade shear modulus
rapidly beyond the elastic threshold strain (0.001%) gave superior predictions under the given conditions. Using such
curves, ultimate bearing capacity predictions produced errors of 0.4% to 10.8%. Using the same input parameters, stresssettlement
predictions were of satisfactory accuracy over the entire settlement range.
Description:
The paper was published in the Proceedings of the 8th
International Symposium on Deformation Characteristics of
Geomaterials ( IS- PORTO 2023) and was edited by
António Viana da Fonseca and Cristiana Ferreira. The
symposium was held from the 3rd to the 6th of
September 2023 in Porto, Portugal.