Date of Award
5-2018
Document Type
Thesis
Degree Name
Master of Science (MS)
Committee Member
Dr. Stephen Moysey, Committee Chair
Committee Member
Scott Brame
Committee Member
Dr. Larry Murdoch
Abstract
The volcanic island Dominica is one of the Windward Islands located near the center of the Lesser Antilles. It contains rugged peaks with steep slopes and has a humid tropical climate. Dominica has four distinct soil types (kandoid, smectoid, allophane latosolic, and allophane podzolic). The distribution of these soils is dictated by climatic factors that influence the degree and depth of leaching. Along with the geology and climate of Dominica, these soils play a large role in determining the prevalence of landslides in different portions of the island. The landslides are typically triggered by high intensity rainfall during tropical storms and hurricanes and since the soil types have varying porosity and permeability, they will react differently to the high rainfall levels. Several recent slope failures were investigated in January 2017 to gain insights into the nature of landslides in tropical settings on island arcs that are experiencing rapid uplift due to their active subduction zone setting. The area selected for investigation is representative of many sites on the island that have roads cutting into slopes at the angle of repose. The roads are often buried and damaged by material from slope failures. The selected slope failure sites are located along a ½ mile section of road and share similar climate, associated soil types, and topographic setting. The two sites selected for further investigations (Forested Site and Lemongrass Site) differ in their vegetative cover (forested vs covered in lemongrass). This investigation analyzed the similarities and differences between the selected sites in order to make a comparison between the failure mechanisms at each site. The Forested Site consists of slope failures CR2 and CR3, while the Lemongrass Site consists of slope failure CR6 and stable slope LG. At these sites, soil cores were collected and taken to the laboratory for soil characterization (plasticity, soil color, and soil name), XRD analysis, and direct shear strength testing. A slope stability analysis was conducted using a limit equilibrium approach. Soil characterization was conducted using the USDA Soil Texturing Field Flow Chart. Results indicate that the two sites are underlain by sandy to silty clay soils with a medium high to high plasticity. XRD analysis revealed a halloysite rich matrix, which is indicative of kandoid soils. Kandoid soils are highly permeable soils found in the oldest volcanic areas and in areas of high rainfall (2100-3750mm) where leaching is moderate but continuous. Drained direct shear tests were completed to determine the effective friction angle and cohesion. The LG sample had a 32 degree friction angle and a value of 128 psf for cohesion. The CR6 sample had a 26 degree friction angle and a value of 159 psf for cohesion. The CR2 sample had a 47 degree friction angle and a value of 320 psf for cohesion. The CR3 sample had a 47 degree friction angle with a value of 0 psf for cohesion after correcting for a calculated negative cohesion which is not possible. These values served as the basis for inputs into a limit equilibrium slope stability analysis using the infinite slope method. Two different types of calculations can be done using the infinite slope method: forward calculations and back calculations. Forward calculations are used to find the factor of safety for slopes that haven’t failed before. For slopes that have failed before, two different back calculations can be done by rearranging the original formula to solve for 1) the friction angle needed for the current slope to stay stable under different conditions, and 2) the height above or below the failure surface that the water table reached when the initial failure occurred. The forward calculations for the slope stability analysis of slope LG at the Lemongrass Site produced a factor of safety greater than 1 which indicates that slope LG is predicted to remain stable. Results of the back calculation analysis determined that at CR2 and CR3 the water table was located an average of 1 ft above the failure surface at the time the slope failed. This means that failure occurred before seepage conditions (seepage conditions occur when the water table is located at the ground surface due to prolonged and heavy rainfall conditions) were fully reached. For slope failure CR6, the water table was calculated at 4.3 ft below the failure surface at the time of failure, which implies that failure occurred before general conditions (general conditions occur when the water table is at the same elevation as the failure surface) were reached. The back calculation analysis also revealed that the slopes must have an effective friction angle of at least 36˚ (measured from horizontal) to remain stable under general conditions (where the water table is at the failure surface) and a friction angle greater than 62˚ for seepage conditions. Slopes with friction angles less than 36˚ will be more likely to fail under general conditions, while those with friction angles less than 62˚ will be more likely to fail under seepage conditions. Since most clays and silty clays have friction angles with maximum values around 30˚, it is anticipated that these slopes would fail under seepage conditions.
Recommended Citation
Fisher, Audrey, "Analysis of Slope Failure on Cochrane Road, Dominica Prior to Hurricane Maria" (2018). All Theses. 2835.
https://tigerprints.clemson.edu/all_theses/2835