A review on modelling of the stress-strain behaviour of soil.

Abstract

Soil has been used as a construction material since antiquity with both success and failure. As the earth material is widely available and relatively economical,it has been found very useful in the construction of foundations, subgrades, embankments and as backfill. The collapsibility of soil is mainly due to the decrease of shear strength and macro-deformation with the increase of moisture content. This can result in some serious problems, such as the differential settlement of the foundation, landslides, and slope instability, resulting in a series of damages of infrastructures and loss of human lives to some degree. In civil engineering, it is very important tomake realistic predictions of the behaviour of soil under various conditions. Triaxial tests are to be carried out to study soil behaviour. The results will indicate the shear damage characteristics during the test. The relation of stress and strain of soils can be expressed using constitutive models. It can provide a brief idea about how soil will behave under different conditions. Application of Neural Network can be used to develop a data-driven constitutive modelling of the soil.