Abstract
Quantitative evaluation for the seismic performance of critical facilities, such as nuclear power plants, to earthquake-induced failure of surrounding slopes, are becoming increasingly important as the level of design for earthquake ground motion for nuclear power plants becomes wider. To carry out earthquake response analysis, the Extended Distinct Element Method (EDEM) is considered ideal. EDEM has uncertainties of initial particle arrangement and shear strength parameters of cohesion parameter and dynamic friction coefficient. Therefore, we investigated the uncertainties in EDEM results of failure timing by initial particle arrangements and cohesion parameters. The purpose of this study was to judge whether the uncertainty of dynamic friction coefficient should be considered when the uncertainty of shear strength between particles in EDEM needs to be considered. The final goal of this research is to build rational EDEM approach in in Probabilistic Risk Assessment (PRA) framework by reducing computational time from ignoring parameters which do not influence on analytical results. We conducted fifty numerical simulations from dynamic friction coefficients and compared these results with those of previous studies on the uncertainty of cohesion parameters. Results show that the uncertainty of dynamic friction coefficients has less influence on the numerical simulation results than cohesion parameters. Therefore, the uncertainty of dynamic friction coefficient may be ignored if the coefficient of variation is somewhat small.
After high magnitude earthquakes such as the Great East Japan Earthquake, the level of the design for earthquake ground motion for nuclear power plants has become wider. Therefore, quantitative evaluations of the seismic performance of critical facilities, such as nuclear power plants to earthquake-induced failure of surrounding slopes are becoming increasingly important as a deterministic approach in regulation. Moreover, evaluation of other aspects other than the design -based earthquake ground motion in Probabilistic Risk Assessment (PRA) is important as the voluntary activity by corporation.
The seismic stability of the surrounding slopes is often determined by the Finite Element Method (FEM) based on ground displacement. For example, the CRIEPI recommends a time history nonlinear analysis for evaluating the stability of slopes, including post-earthquake residual displacement, and predicting a failure range when large deformations and displacements occur (Ishimaru, 2017).
