Model development to capture the improvement of shear strength of soil using Australian native vegetation

The use of vegetation for ground improvement is a sustainable, environmental friendly and cost effective approach. For railway corridors, this technique is now increasingly looked at for improving the shear strength and stiffness of subgrade soil apart from obvious environmental benefits attributed to wind barrier controls, and for reducing the effects of greenhouse gases. The increase in soil shear strength and stiffness is mainly due to the suction induced by root water uptake and the mechanical reinforcing effect provided by the tree roots. This doctoral research mainly focuses on investigating the integrated behaviour of suction and root reinforcement in shear strength improvement of soil while past researches have considered these two aspects separately as independent components. The most rational way of capturing the true behaviour of vegetated ground is to treat the geo-hydraulic and mechanical properties as an integrated system, in view of the fact that root-permeated soil often remains in unsaturated condition due to the continual climatic process of evapo-transpiration. A series of laboratory and field investigations was carried out to examine the behaviour of a suction-reinforcement integrated system, and accordingly, a mathematical model was developed to support the experimental observations. A MATLAB simulation was also carried out based on the governing equations developed herein. The effect of coupled suction-reinforcement approach on the increased shear strength and the potential root failure modes were identified through direct shear testing. The theoretical predictions were found to be in good agreement with the laboratory results. Furthermore, the results obtained from a field investigation conducted on a site located at the University of Wollongong campus verified the intricate relationships between the root water potential and measured matric suction variations. A two-dimensional finite element analysis (plane strain) was carried out using PLAXIS-2D to simulate and demonstrate the native vegetation process in a practical application, in which the complex 3D root system was simplified to a 2D approximation. The field results reported by others (e.g. Potter, 2005 and Fatahi, 2007) were used in a finite element analysis and then the results of the initial settlement between non- vegetated and vegetated ground were compared to the increments of suction. The increase in soil shear strength along a ‘green’ rail corridor was captured in this PLAXIS simulation, further supported by MATLAB analysis based on the writer’s mathematical model. In particular, the hardening soil model available in PLAXIS was adopted for the root-permeated section, and the application and reliability of the model could be further validated by simulating the direct shearing process with and without root reinforcement.The FEM analysis indicates that the initial settlement of rail corridors with vegetation can be as much as 50% less than that of non-vegetated ground, and this benefit is further accrued with the increase of suction generated in the soil.