Centrifuge modelling of hydrological and mechanical effects of vegetation on slope stability

Abstract

The use of vegetation to stabilise slopes from shallow failure has been increasingly attractive from the environmentally friendly perspective. Previous research mainly focused on mechanical effect (i.e., root reinforcement) of vegetation on slope stability, while the hydrological effect (i.e., water uptake (or transpiration) induced negative pore-water pressure (PWP)) is less studied. The objective of this research is to investigate effects of vegetation on slope stability. To achieve this, a novel technique capable of modelling both mechanical and hydrological effects of vegetation in geotechnical centrifuge was developed. The technique involves using a branch cutting to model the plant root to which vacuum is delivered to generate a pressure gradient. This enables water to be extracted from the soil into the branch cutting and thus a negative PWP to be induced in the surrounding soil. This technique was implemented in two series of centrifuge tests (rainfall test and rising g-level test) to investigate: (i) retention characteristics of the induced negative PWP subjected to the infiltration of rainfall with different durations; (ii) effect of antecedent water uptake duration on wetting response of the induced negative PWP; and (iii) failure characteristics of vegetated slopes. Transient seepage analyses were carried out to further interpret results obtained from centrifuge tests. It is found that negative PWP can be induced by the newly developed technique and the PWP distribution is reasonably consistent with field measurements of vegetated slopes. After a one-hour 108 mm/hr rainfall (20-year return period), at least -5 kPa PWP can be uniformly maintained within 2.5 times root depth. Given that rainfall duration, there exists a threshold rainfall intensity (135 mm/hr), higher than which would not help in further destroying the negative PWP as the soil has reached its infiltration capacity. Once the rainfall duration further extends, however, the negative PWP is destroyed significantly as more water is infiltrated in. On the other hand, regardless of antecedent drying durations, the effect of water uptake can reach 2.5 times root depth. While upon rainfall with an extended duration (i.e., 2.5 hrs), long pre-drying duration does not contribute to maintaining the negative PWP, which implies that hydrological effect of vegetation vanishes. Stability analysis of the model slopes shown that compared with the mechanical contribution of vegetation to factor of safety, hydrological effect (once negative PWP maintained) is more pronounced. Moreover, due to the presence of negative PWP, critical slip surface goes deeper. Unlike the conventionally assumed pullout failure in literature in assessing the stability of vegetated slopes, failure was observed in the rising g-level test while the roots almost remain intact. The failure initiated at slope toe, and slip surfaces were generally shallow and parallel to slope surface. Due to the shallow slip surface depth (30-45% of root depth), the confining stresses around roots in the sliding zone are low and do not provide enough frictional force to pull the embedded roots out.

Date of Award	2015
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology