Voxelization approach to model soil reinforcement by root systems: application for landslides risk assessment of forested hillslopes

English Abstract:

Root reinforcement is considered as an increase of soil shear strength and this help in reducing slope failures. To investigate effects of tree root systems on slope stability, field experiments can be conducted. Such tests are difficult to perform. Numerical model appears as an alternative approach to investigate root reinforcement and risk assessment of forested slopes without taking explicitly into account roots elements in soil.

This paper presents a new and an integrated model of soil reinforcement. The model is based on voxelization technique which consists in converting 3D geometric objects from continuous geometric representation into a set of voxels that best approximates each element of a root system. Based on this technique, a root system located in a block of soil can be divided into a set of voxels or volume elements. In each voxel, roots parameters such as root area ratio, tensile resistance and roots mean orientation can be extracted for the purpose of local soil additional cohesion computation. Root cohesion in different voxels can therefore be assigned to a similar domain meshed into 3D finite elements for the field of strain computation. The process enables to search finite elements at the same spatial position as the considered voxel prior to root cohesion assignment. With this model, simulations of forested slope failure and direct shear tests of rooted-soil were performed in a finite elements environment. Results of the simulations are found to be in good accordance with similar experimental tests found in the literature. This help to investigate slope reinforcement with different types of root systems and plantation patterns. The investigations showed that, among the three main families of roots architectures (tap, heart and plate root systems), the tap root system is found to be the best in terms of soil reinforcement. Staggered in rows forest pattern was found to increase slope stability much better than all other patterns investigated.