The Progressive Development and Post-failure Behaviour of Deep-seated Landslide Complexes

Abstract

The concept of progressive shear surface development remains to be fully understood. This study aims to investigate the mechanisms of progressive shear surface development to failure, and the subsequent post-failure behaviour, of deep-seated landslide complexes.

Field samples, collected from the Ventnor deep-seated landslide complex, Isle of Wight have been tested in a series of bespoke triaxial and back pressured shear box tests, in which field failure conditions are simulated by elevating pore water pressures under a constant deviator stress. These laboratory simulations have been calibrated with a detailed analysis of ground movement, groundwater and antecedent rainfall relationships from real-time site monitoring records from the site.

A model to explain the progressive failure of landslides in cohesive materials is proposed. The model demonstrates that pre-failure movement in landslides occurs during the progressive shear surface development through the development of micro-cracks within the slope. This process is active during periods when porewater pressures are sufficiently elevated in the slope. These micro-cracks eventually lengthen and coalesce as stresses concentrate at micro-crack tips, leading to the development of a singular shear surface which is observed as tertiary creep. The study confirms that the onset of ‘Saito linearity’ in 1/v – t space is associated with this tertiary creep phase. This mechanism can occur through time dependent creep at constant stresses from below the peak strength envelope. The results have been validated against real-world landslide monitoring data to provide a new understanding of the shear surface mechanisms operating within the Ventnor landslide.

The results contribute new knowledge regarding the mechanisms of shear surface development and provide an improved understanding of these mechanisms by developing one of the first high quality data sets that combines laboratory and field data. Finally, the study provides a new method to inform future landslide behaviour predictions for deep-seated slope failures.