Identifier

etd-10242005-122322

Degree

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

Document Type

Dissertation

Abstract

A non-invasive experimental method and a clustering DEM model were developed in this study to investigate micro-macro behaviors of real granular materials with irregular particle shape configurations. The investigated behaviors include global deformations, failure strengths and residual strengths, stress and strain distributions, local coordination number, local void ratio, particle kinematics, and fabric orientation distributions. The experimental method includes an approach to automatically identify and recognize multiple particles using x-ray computed tomography imaging (XCT) and an enhanced approach to digitally represent microstructures of granular materials. The digitally represented microstructure can be directly employed for numerical simulation setup. A compression test and a direct shear test on coarse aggregates were conducted and analyzed using this method. The experimental measurements were applied for the evaluation of DEM simulations. The clustering DEM model provided in this study extends the conventional DEM model by incorporating actual microstructure of materials into simulations. Real irregular particles were represented by clusters of balls in the clustering DEM model while spherical particles were employed in the conventional DEM model. The PFC3D commercial software was applied for 3D DEM simulations of the compression test and the direct shear test. Compared to the conventional DEM model, the clustering DEM model demonstrated a better capability of predicting both the micro and macro behaviors of granular materials, including dilation, strength, particle kinematics, and fabric evolution. Fabric distribution was investigated for both the conventional DEM model and the clustering DEM model. The clustering DEM model described the fabric distribution of actual materials more precisely. This feature enabled it to simulate micro-macro behaviors of materials more accurately. A theoretic stress-fabric tensor relationship was also evaluated using the simulated stress and fabric distributions based on the actual microstructure of a real material. This relationship incorporated the anisotropic microstructure characteristics of the material. Whether it can better describe behaviors of granular materials was evaluated in this study. Generally, this research provides a more inherent understanding of granular materials through both DEM simulations and experimental validations.

Date

2005

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Linbing Wang

Share

COinS