Simple-shear and direct-shear behaviours of grass-rooted soils

Root reinforcement to soil has been commonly quantified by direct-shear tests (DSTs). Despite the simplicity and popularity, this testing method is often criticised for imposing a predefined shear plane, unrealistically representing the in situ stress conditions. Suitability of using DST to characterise the mechanical properties of rooted soils has never been discussed before. This study aims to investigate the differences in the shearing behaviour and stress–dilatancy relationships of rooted soils obtained from simple-shear tests (SSTs) and DSTs. A new stress–dilatancy relationship for rooted soils was derived to explain the additional soil dilatancy contributed by roots upon shearing. The measurements and prediction consistently revealed that the DSTs substantially overestimated the cohesion (by 100%) and peak friction angle (by 15%) of rooted soils, compared to SSTs. This phenomenon is due to the much greater principal stress rotation in the DSTs due to the undesirable stress concentration near the forced horizontal shear plane, causing virtual increases in root reinforcement. Cautions should be taken when using DST to quantity root reinforcement to prevent unsafe engineering design. Predictions made by the validated model showed that the effects of root-induced dilatancy and the associated increase in peak friction angle were more prominent for shallow rooted soils.