Abstract

Arriving at a realistic model is of importance in foundation analysis by the extreme difficulty of modeling the soil-structure interaction. To this purpose, a plenty of models have been introduced, over the years, range from simple single parameter models to complex mixed models. One of the most popular and simple methods, which is considered as “one-parameter model”, is the Winkler’s model. In this model, the soil continuum is superseded by an infinite number of closely spaced springs, whose stiffness describes the behavior of the elastic foundation. Replacing the soil with a group of springs leads to the model discontinuity and uniform settlement profile of footing, as, every single spring is only responsible for its own vertical load and is not affected by the shear stress or the deformation of the adjacent springs. Two main inborn imperfections of the Winkler’s model include: the necessity of determination of the subgrade reaction modulus (K s ), and uniform settlement of the foundation under distributed load. Actually, the second one is in opposition to the dish-shaped settlement profile of the same footing when considering the soil as continuum using Finite Element Method (FEM). The modulus of subgrade reaction is not a soil property and could have a different value for a specific footing owing to the natural variations in soil properties. The footing dimensions and rigidity, load level, soil condition, and soil model have a direct impact on K s . In the first part of paper, the effect of foundation rigidity on the K s distribution is investigated. Moreover, the commonly-used correlations of K s estimation, i.e. Biot, Vesic and elastic-based correlations, are examined to find the most accurate one, using FEM. In order to amend the Winkler’s model concerning the latter weakness, the foundation is divided into central part as well as side parts and assigning different subgrade modulus to the parts, e.g. a value of K s to the center part and double to the side parts. The settlement profile of the footing with various side-part widths are compared with the corresponding FEM model to find the best case. The results show that with this dividing technique the desired dish-shaped settlement profile could be achieved by error less than 20%.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call