Download files
Abstract
The water retention curve (WRC) is a fundamental property of porous media that has extensive applications in different science and engineering disciplines. Past experimental evidence has shown that the WRC evolves with structural changes in soil matrix. This study is aimed to address the shortcomings and ambiguity issues in the previous experimental and modelling works with respect to the WRC for deformable soils.
In the experimental program of the study, in addition to the baseline experiments, true constant volume WRCs were obtained in laboratory for the first time in the literature. Based on this data, the evolution of the WRC and the embedded parameters with void ratio were analysed in detail.
In addition, pitfalls in the common approaches to obtaining material parameters from WRC, particularly the air entry value (AEV), for deformable soils are discussed in this study, and a consistent graphical approach for the determination of AEV based on an understanding of the effects of stress history and volume change on gravimetric water retention behaviour is presented.
This work also presents two modelling approaches to account for the volume change dependency of the WRC in a deformable soil. The first approach is based on energy considerations. The model does not require any additional material parameter apart from the parameters specifying the WRC for the reference volumetric strain and incorporates the effects of hydraulic hysteresis and volume change dependency of the scanning curve which is rarely addressed in the literature. A unique feature of the model is its ability to capture the change in the hydraulic path of the soil from scanning to main in a mechanical loading event.
A fractal-based volume dependent WRC model has also been developed in this study. A feature of this model is that it relates the WRC parameters to easily quantifiable parameters. Moreover, the fractal dimension of pore size distribution varies with void ratio in the presented model, an aspect frequently neglected in the literature. The comparison of results from the two models that were developed independently in this study shows that the models generally lead to consistent predictions of the WRC evolution with void ratio. The models provide a basis for correlating the effective stress parameter to the soil textural properties.
The application of the models presented in this study is also demonstrated through an extensive comparison between numerical predictions and experimental data from this study and other published works. Very good agreement between model predictions and experimental results is observed highlighting the ability of the proposed models in capturing the volume change dependency of the WRC. These contributions have to date resulted in publication of three journal papers and five conference papers.