Movement of reactivated landslides

Slope movements cause considerable damage to life and property. While slow moving landslides do not typically cause loss of life, the long-term continous movements can severely damage structures or interrupt the serviceability of highways, railways, and pipelines, requiring continous maintenance. The primary objective of this thesis is to examine the post-failure/reactivated deformation behavior of landslides in stiff clay and shale slopes in order to establish typical baseline displacement rates. A literature review, back-analysis of landslide case histories, and laboratory testing program were conducted to improve understanding of general characteristics and behavior of post-failure movement of landslides.To study the displacement rates in slow-moving reactivated landslides, a new laboratory constant normal- and shear-load direct shear apparatus was designed and constructed. A series of tests was carried out using reconstituted pre-cut specimens. In order to specify shear stress level for the load-controlled tests, a series of deformation controlled reversal direct shear and ring shear tests was carried out for the measurement of residual shear strength. The laboratory testing program also included Atterberg limits, clay-size fraction, as well as scanning electron micrograph and x-ray diffraction analysis of the materials tested. For the pre-cut surfaces used in this study, shearing displacements on the order of 5 mm was sufficient to bring these surfaces to residual condition. Data for laboratory tests reported in the literature are examined to extract additional data on displacements required to reach the residual condition for different intact and pre-cut specimens.For shear surfaces at residual condition creep deformations are observed at ail shear stress levels. At moderate levels of shear stress, creep displacements continued for relatively long periods of time (more than 1-3 days in the current tests), with approximately constant creep displacement rates in the range of 3 to 10 mm/year. An increase in shear stress level results in an increased rate of creep. It is observed that, at high shear stress levels, creep displacement initially occurs at a decreasing rate until a minimum creep displacement rate is reached, and then the displacement rate increases, resulting in eventual acceleration to failure. At any shear stress level, the logarithm of the displacement rate decreases linearly with the logarithm of time, until at high shear stress levels a minimum rate is reached and is followed by rapid increase in displacement rate. The slope of the linear relationship is more or less independent of the shear stress level, and is in the range of 0.78 to 1.18 for load increments that do not lead to failure in about 1-day measurement period, and 0.33 to 0.70 for the linear portion of the load increments that lead to acceleration and failure within a 1-day observation period. The minimum displacement rate immediately before acceleration to failure is observed to be inversely proportional to remaining time to failure.Failure time prediction methods are investigated based on the laboratory creep movements leading to accelerated displacements. Prediction of failure time using (a) Saito's time to failure with minimum creep displacement rate, (b) Fukuzono's inverse velocity, (c) Picarelli et al.'s acceleration plot with time to failure, are examined with the creep displacement test data obtained in this study. Although general conclusions cannot be reached based on the limited number of tests and, under limited time period (about 1 day) of creep displacement observation, these failure time prediction methods show promise for estimating failure time for landslides.Because the displacement rate of "very slow" and "extremely slow" moving reactivated landslides in stiff clays and clay shales, is typically controlled by fluctuations in piezometric levels the relations between rate of movement and piezometric levels and factor of safety (proportional to inverse of shear stress level) are investigated. For the landslides investigated in this study there appears to be a nonlinear (power) relationship between the rate of movement and the factor of safety of reactivated landslides. It is concluded that for an approximately 0.1 increase in factor of safety (10% increase) the rate of movements decrease by about 5 to 20-times. (Abstract shortened by UMI.)...