Macro- and Micro-Scale Effects of Pluviation Based Sample Preparation in DEM

Abstract

Previous laboratory studies and Distinct Element Method (DEM) simulations have shown that the preparation method used to create samples for testing can affect the sample fabric and hence the mechanical response observed. The specimen preparation methods traditionally used in the laboratory were developed to recreate a given depositional pattern or in-situ soil fabric. Differences in soil response have been attributed to variations in the initial fabric of the sample even when the samples are created at the same densities. If DEM simulations are to provide meaningful insight into soil response observed in element tests, it is important for the initial state (packing density and stress level) and fabric anisotropy of the computer generated sample to closely match the physical reality. This is particularly true where element tests are used to validate DEM models. A key challenge is the lack of quantitative data on the fabric of real physical test specimens. Several approaches have been documented in the literature for generating the initial configuration of specimens for DEM simulations, i.e. artificially creating a percolating (stress transmitting) granular material. Approaches that are commonly used include radius expansion, compression using rigid boundaries, and pluviation under gravity loading. This paper examines the influence of the method chosen on the macro- and micro-scale properties of the material generated. The methods considered involved generating particles as a diffuse cloud at their target size, followed by pluviation under gravity loading with and without a mesh that modeled a sieve. The study fits within a broader research project in which multidirectional simple shear tests on steel spheres will be replicated in DEM simulations. © 2012 American Society of Civil Engineers.