Physical and numerical investigation of a cemented granular assembly of steel spheres

Abstract

SUMMARYIn this paper, steel spheres embedded in a cement matrix were studied using numerical and physical ISRM testing procedures. A challenge in discrete element simulations is to select appropriate micro-mechanical models and parameters, to recover the observed macro-mechanical behavior. An ideal experiment on cohesive granular assemblies constructed identical to numerical ones would validate these micro models for a set of measured micro-parameters. The first part of the paper summarizes the previous studies in this area, outlines such experimental methodology and depicts the steps followed for the preparation and the testing of cemented granular assemblies together with the derivation of micro-parameters. The second part discusses the results of numerical and physical ISRM standard tests including uniaxial and triaxial compression, Brazilian tensile and shear box tests. Physical samples were prepared using steel balls bonded with Portland cement, cured under controlled laboratory conditions and tested in compression, tension and shearing. Acoustic emissions were monitored in uniaxial tests to characterize the damage thresholds relative to volumetric strains. Numerical simulations were conducted with PFC 3D using micro-mechanical parameters derived from physical testing. Parametric sensitivity studies were carried out to look into the dependency of macroscopic responses on the parameters. The results from both numerical and physical tests showed good correspondence in macroscopic behavior i.e. peak strength, stages of damage, mode of failures. However, the numerical simulations reflected a stiffer mechanical response than physical assemblies.

Characterization of clastic rock using a biconcave bond model of DEM

“…In Figure , as the average particle radius ( r avg ) increases, the modulus E 50 and strength σ c decrease, but r avg is independent of ν 50 , revealing that r avg has no effect on lateral deformation. These findings are consistent with those reported previously . For the parallel bond model, the particle radius ( r avg ) almost has no effect on E 50 and ν 50 , and greater particle size also induces lower strength.…”

A bond failure criterion for DEM simulations of cemented geomaterials considering variable bond thickness

“…Experiments to directly testify the existing bond contact models and to develop rational bond failure criteria instead of analytical approaches have been tried. The load–displacement relations at the grain scale have been experimentally attained from analog soil particles or uncemented nature soil/rock particles subjected to simple loads . To test real cemented soil grains subjected to a variety of loading conditions remains challenging because of the extreme difficulty in duplicating loading paths and the geometry of the bonds with necessary precision.…”

“…As a result, analog soil grains are used in this study to enable reproducible experiments that give rise to a reasonably simple contact law for DEM codes. The concept of analog soil grains is not new, and relevant work using aluminum rods , steel spheres , and glass ballotini among others can be found in the literature. This study idealizes the 3D cemented soil grains as glued aluminum rods to imitate a two‐dimensional (2D) plane strain condition, which can be directly implemented into a 2D DEM.…”

A modified particle contact model for matching the ratios of uniaxial compressive to tensile strength of brittle rocks

“…Moreover, the clumping model displays ductile post-peak behavior, which is inappropriate for brittle rocks. The use of particles with intricate shapes complicates the correlation between numerical simulations and physical materials (Akram and Sharrock 2010). On the premise of not significantly reducing the calculation efficiency, the number of the bond can be increased by expanding the interaction range between particles, and the interlocking effect can be indirectly strengthened.…”