Experimental and numerical study of the soil arching effect using the Discrete Element Method

Abstract

In the present article, the “arching effect” [R. Handy, 1985] is reproduced by modeling the Terzaghi’sTrapdoor test with the Discrete Element Method (DEM). The numerical model results will be compared with ex-perimental well instrumented tests from previous stages of this research [G. Pardo, 2013]. The adopted modelingapproach describes the soil as an assembly of particles, reproducing the macroscopic soil behavior from micro-mechanics interactions between individual particles [P. Cundall, 2000]. The calibration of micro-mechanics param-eters is performed from direct shear soil tests, using an arrangement of particles with sizes consistent with the realgranulometry and void ratios similar to those obtained experimentally. The calibrated set of parameters is used toreproduce the trapdoor experiment. The goal is to accurately reproduce the trapdoor test, in order to characterizeload transfer mechanisms and shear strain concentrations that naturally occurs when parts of the support system ofthe soil starts yielding or moving, while the rest stays stationary.