Projects

Background

Most of the earth’s surface in arid and semi-arid zones is formed by unsaturated soils. Unsaturated soils usually exhibit more complex behaviour compared to saturated and dry soils because of the presence of matric suction inside their skeleton. Matric suction can influence several aspects of soil behaviour such as the rolling and shearing resistance of particles in granular soils and the volume changes of soils under monotonic and hydro-mechanical cyclic loading. In addition to matric suction, the behaviour of unsaturated soils can vary with soil types and factors such as the evolution of stress-induced soil anisotropy. Because of the differences that sands and clay may show to external loads, it is commonly observed that loose sand can be prone to liquefaction under undrained conditions, however, this scenario may not happen to clay. Characterisation and unification of the behaviour of unsaturated sands and clays can enhance the accuracy of design in many engineering problems associated with unsaturated soil response.

Several constitutive models have been developed up to now and can predict some essential characteristics of the behaviour of unsaturated soil, such as wetting-induced volumetric collapse. Although many breakthroughs have been achieved, several challenges remain. One of the challenges is the precise description of soil behaviour under cyclic loading. Several models successfully developed from the so-called Cam-Clay model and assume soil behaviour is purely elastic for all stress paths prior to reaching the yield surface. Since plastic deformations can significantly accumulate in the case of cyclic loading even prior to reaching a yield surface, the simulation of soil cyclic behaviour is beyond the capability of these models. To address this issue, one of the well-accepted methods is multi-surface plasticity theory, however, there is a need to develop novel computational algorithms (e.g., for the prevention of over-shooting phenomena) to enhance the performance of this class of models in numerical simulations. Another challenge is the unification of the response of sands and clays using a single set of parameters due to the differences in their microstructure. Despite the attempts to model various types of saturated soils in a unified manner, a less similar effort has been directed to unsaturated soils.

This research aims to rectify the shortcomings of earlier studies by establishing a unified constitutive model for unsaturated sands and clays. One of the promising features of the research is the inclusion of the effect of stress-induced anisotropy in the description of plastic deformations. The inclusion of this feature allows the proposed model to give an enhanced description of the behaviour of anisotropically consolidated (K0-consolidated) soils (which are frequently encountered in practice) to cyclic loading. The project also aims to propose a novel and robust computational algorithm that minimises the problem of overshooting and allows the use of the model in solving complex boundary-value problems involving the cyclic response of unsaturated soils.

The proposed model will also use the MPK framework. The MPK framework facilitates the development of a constitutive modelling technique, in which commonly measured compaction curves is directly used in the calibration process. It is also possible to treat suction as a dependent variable in this framework, where suction can be predicted by coupling water retention behaviour.

The research would provide a deeper insight into the mechanics of unsaturated soil under cyclic loading conditions and allow the simulation of both sands and clays in a unified manner. This unified nature can significantly facilitate the model application in the design and construction of civil structures, particularly pavements, that predominantly operate under unsaturated conditions.

Project objectives

1. Propose a unified constitutive model for unsaturated sand and clay that allows the description of the cyclic and monotonic response of these types of soils by using a single set of parameters.

2. Further refine and verify the proposed model based on the targeted experimentation.

3. implement the proposed model into a finite element package to study a series of engineering problems associated with the unsaturated soil under hydro-mechanical cyclic loading.