An Implementation of probabilistic seismic hazard analysis in selection of seismic coefficient for pseudostatic analysis of slope stability


Thesis Type: Post Graduate

Institution Of The Thesis: Middle East Technical University, Turkey

Approval Date: 2017

Thesis Language: English

Student: Kübra Gedikaslan

Co-Consultant: MUSTAFA TOLGA YILMAZ, MURAT ALTUĞ ERBERİK

Abstract:

The seismic coefficient, kh, is a seismic design parameter. This parameter defines the ratio of inertial force acting on a mass to its weight, and is practically used in pseudostatic analyses of seismic slope stability. This design parameter can be assigned by the principles of performance-based design after simplifications regarding the relationship between ground displacement and probabilistic seismic hazard. A widely used simplification is to consider Newmark's sliding block analogy for estimations of seismic displacement on sloping ground. Therefore, the permanent sliding-block displacements (D) corresponding to a set of critical accelerations were computed for a sample of 70 horizontal-motion accelerograms. A prediction equation was developed by using the method of least squares. Moment magnitude, Joyner and Boore distance, style of faulting, average shear-wave velocity in top 30m were chosen as the parameters necessary for prediction of D. A probabilistic seismic hazard analysis was performed by implementing this prediction equation. A zone near Şebinkarahisar district of Giresun province was chosen as a case study. The results were used for computation of sliding (critical) accelerations corresponding to the displacements of 1 cm, 10 cm and 100 cm. The return periods of the random events that these displacement limits are exceeded were calculated. The critical acceleration corresponding to a particular displacement threshold, related to desired performances of facilities near to sloping ground, and to a particular return period, related to intensity of ground motion to be considered in design, was supposed to be appropriate parameters for final selection of kh. Hence, the spatial distributions of kh on the study zone were shown. The ranges of kh were compared with the topographical and geotechnical characteristics of the area. A criticism on the prediction equation for D was presented by putting emphasize on the variance of prediction and on the site effects.