The seudostatic method of dam stability: a critical analysis

Abstract

A 200 m high tailings dam located in central Chile was seismically evaluated in four different cross sections using a 2D finite element model. To perform the analyses two acceleration records were used, one from the dam seismic risk study and another one from the 27^th February 2010 Chile earthquake. In parallel, a number of seudostatic analyses were carried out applying a horizontal acceleration (defined in terms of a global seismic coefficient) to the same finite element model used in the dynamic analysis in order to reproduce the deformation field created by the later. A method to compare the deformation fields between the two models was developed. In this manner, the seismic coefficients that better approach the field deformations from both types of analyses were estimated. It is observed that in most cases the Chilean state-of-practice largely over-estimates the selected values of the seismic coefficient. Moreover, for different dam heights and/or different seismic frequencies, significant differences between the seismic coefficient values have been found. This study presents a first approach with the aim ofpointing out the importance of the adequate selection of seismic coefficient values used in limit equilibrium analyses.