Date of Award

8-2008

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Civil Engineering

Advisor

Juang, Charng H

Committee Member

Amirkhanian , Serji N

Committee Member

Ravichandran , Nadarajah

Committee Member

Schiff , Scott D

Abstract

Assessment of slope stability or landslide is a complex geotechnical problem that involves much uncertainty. To consider combined effects of the uncertainties in soil parameters and changes in pore water pressure, reliability theory may be employed. In this dissertation study, a reliability-based framework for the probabilistic analysis of recurrent landslides is developed.
Generally, the stability of a slope is analyzed deterministically with methods of limit equilibrium and a factor of safety is computed. This approach is referred to as the deterministic approach. However, a deterministic analysis alone may not be sufficient in assessing the risk level of a landslide in view of the uncertainties inherent in the input parameters.
Eliminating uncertainty from a landslide site is difficult due to the large variation of the groundwater level and uncertainty in the shear strength parameters. However, a reliability-based approach, which considers the variation in the ground water level and uncertainty in the shear strength parameters, is shown to be useful in coping with the uncertainties. With this approach, different levels of warning for the landslide stability are developed. Furthermore, the relationships between increased pore-water pressure and total rainfall are demonstrated to be a useful tool for assessing slope instability for similar landslide-prone areas under similar climatic conditions and geologic settings.
Using the probability of failure calculated with the reliability analysis, a strategy is implemented for predicting the risk of an impending landslide and a cost assessment is performed to evaluate the feasibility of the alternative engineering decisions for remediating the landslide.
An important landslide which occurred in Tegucigalpa, Honduras in October 1998 provides a case study to demonstrate the proposed framework. Whereas, previous studies have attempted to define the potential danger of the El Berrinche landslide, uncertainty has not been adequately addressed, which compromises realistic projections of the failure. To date, no consensus among local authorities exists on what solution is feasible and cost-effective. The landslide at El Berrinche involves a complex failure scheme. Two major types of failure have been identified: a small debris flow which started at the toe of the landslide, and a near-rotational failure along a relatively thin shear plane area which occurs within the altered shale. The results of this case study demonstrate the advantages of employing the probabilistic approach over the use of the deterministic approach in a risk-based engineering decision making regarding the alternative remedial measures.

Share

COinS