An analytical solution for geotextile-wrapped soil based on insights from DEM analysis

Abstract

This paper presents a novel analytical solution for geotextile-wrapped soil based on a comprehensive numerical analysis conducted using the discrete element method (DEM). By examining the soil-geotextile interface friction, principal stress distribution, and stress-strain relations of the constituent soil and geotextile in the DEM analysis, a complete picture of the mechanical characterization of geotextile-wrapped soil under uniaxial compression is first provided. With these new insights, key assumptions are verified and developed for the proposed analytical solution. In the DEM analysis, a near-failure state line that predicts stress ratios relative to the maximums at failure with respect to deviatoric strain is uniquely identified; dilation rates are found to be related to stress ratios via a single linear correlation regardless of the tensile stiffness of the geotextile. From these new findings, the assumptions on the stress-state evolution and the stress-dilatancy relation are developed accordingly, and the wrapped granular soil can therefore be modeled as a Mohr-Coulomb elastoplastic solid with evolving stress ratio and dilation rate. The development of the proposed analytical model also demonstrates an innovative approach to take advantage of multiscale insights for the analytical modeling of complex geomaterials. The analytical model is validated with the DEM simulation results of geotextile-wrapped soil under uniaxial compression, considering a wide range of geotextile tensile stiffnesses. To further examine the predictive capacity of the analytical model, the stress-strain response under triaxial compression conditions is solved analytically, taking both different confining pressures and geotextile tensile stiffnesses into account. Good agreement is obtained between the analytical and DEM solutions, which suggests that the key assumptions developed in the uniaxial compression conditions also remain valid for triaxial compression conditions.

Probabilistic calibration of discrete element simulations using the sequential quasi-Monte Carlo filter

“…The implementation of the rolling/twisting governed by CL and HM contact laws, at the 60th data assimilation step scheme for the CL contact law, on the other hand, calculates the rolling stiffness from the radii of the contacting grains, the tangential contact stiffness and dimensionless factor β m . Nevertheless, the agreement between the posterior PDFs using the same contact law is generally good, and in line with most large-scale DEM simulations that use scaled grain size distributions to represent granular soils [6,14,20]. The agreement of the posterior distributions in Fig.…”

“…From just a few micromechanical parameters, DEM can provide comprehensive cross-scale insights [2,6,15,42] that are difficult to obtain in either stateof-the-art experiments or sophisticated continuum models. However, automated and systematic calibration of these parameters against macroscopic experimental measurements is still lacking, and often takes significant effort from DEM analysts.…”

Hydro‐micromechanical modeling of wave propagation in saturated granular crystals

“…Comparative studies have shown that the Noble‐Torczynski method causes numerical dissipation, with the accuracy depending on the relaxation time, whereas the MEM is prone to high‐frequency fluctuation because of the binary transition between solid and fluid nodes. Nevertheless, the noise appears at scales much smaller than particle sizes and thus is smeared out after local averaging . For this reason, the MEM has been selected as the fluid‐solid coupling scheme for this work.…”

“…Nevertheless, the noise appears at scales much smaller than particle sizes and thus is smeared out after local averaging. [50][51][52][53] For this reason, the MEM has been selected as the fluid-solid coupling scheme for this work. Fluid advecting onto the surface of a solid particle is simply bounced back half way between the solid and the fluid nodes by setting…”

“…LB-DEM simulations using the momentum exchange method (MEM) is prone to high-frequency fluctuation because of the binary transition between the fluid and solid nodes. Nevertheless, the noises appear at a spatial scale much smaller than particle sizes and thus will be smeared out after performing local volume averaging [36,37,38,39]. It is expected that as long as the resolution of the solid particles is sufficiently high, the high-frequency noises will not affect the pore-scale flow behavior.…”