Mesoscopic representation of conventional concrete and rock-filled concrete: A novel FEM-SBFEM coupled approach
Weichi Xu,Yuande Zhou,Yutai Guo,Feng Jin
DOI: https://doi.org/10.1016/j.compgeo.2024.106820
IF: 5.218
2024-10-13
Computers and Geotechnics
Abstract:A thorough characterization of the mesostructure of concrete serves as a fundamental cornerstone for investigating its complex mechanical response at the mesoscale. A coupled FEM-SBFEM (Finite element method − scaled boundary finite element method) model is developed for mesoscopic modeling of conventional concrete (CC) and rock-filled concrete (RFC). This model incorporates a novel RAM (Random Aggregate Model) generation procedure based on Laguerre tessellation, allowing for the construction of coarse polyhedral aggregates with diverse grading schemes and adjustable aggregate volume fractions. Moreover, a framework has been developed for the automatic generation of prelaid rock skeletons, which accurately encapsulate the distinctive self-sustaining skeletal characteristics inherent to RFC. In a departure from conventional FEM, the SBFEM in this approach discretizes each coarse aggregate using a singular polyhedral element, resulting in a significant reduction in degrees of freedom. The proposed mesoscopic construction method is adopted for the prediction of elastic properties for both CC and RFC. Numerical samples of 48 CC specimens and 13 RFC specimens, with various aggregate volume fractions and rockfill ratios, are constructed using Monte Carlo simulations, and the results are compared with experimental and numerical data in literature. Statistical analyses are performed to investigate the impacts of aggregate volume fraction and anisotropic behavior on the elastic properties of CC and RFC. The results demonstrate that RFC exhibited an elastic modulus approximately 7.32 % higher than CC at the same coarse aggregate volume fractions. Furthermore, RFC exhibits a more substantial degree of anisotropy than CC. The proposed FEM-SBFEM coupled approach presents the capability to accurately predict the elastic behavior of concrete materials, and can be extended for a comprehensive investigation of the linear and nonlinear properties of actual RFC that comprises extremely coarse aggregates.
computer science, interdisciplinary applications,geosciences, multidisciplinary,engineering, geological