DEM simulations on the influence of carbonate precipitation on liquefaction mitigation of sand

Abstract

Microbially induced carbonate precipitation (MICP) is an effective method to mitigate liquefaction of sands under cyclic shear loading. In this paper, discrete element method (DEM)-based simulations of the cyclic direct simple shear (CyDSS) tests are carried out using a three-dimensional shear box geometry to evaluate the liquefaction resistance of untreated, MICP-cemented, and reconstituted sands. A six-parameter contact model previously used for undrained and drained triaxial tests on sand is used in the DEM analysis. Verification and validation of the DEM model is implemented through modeling of experimental CyDSS results on loose Fraser River sand at several values of cyclic stress ratios (CSR). The DEM model parameters for untreated and microbially treated Ottawa 20–30 sand are obtained by matching the simulated deviatoric stress and excess pore pressure from strain-controlled undrained isotopically consolidated triaxial compression tests. The DEM model is shown to be capable of accurately representing the cyclic strength curves for untreated, treated, and treated but reconstituted sand using parameters derived from the triaxial test data. The simulated stress path, shear strain evolution, and the evolution of coordination number are used to further examine the effect of MICP on liquefaction in CyDSS tests. The DEM model can be used to design cementation levels and develop cyclic strength curves for different types of sands to aid engineers in designing liquefaction mitigation strategies using MICP.