Insight into the Inherent Randomness of Concrete Properties Using the Stochastic Micromechanics

Qing Chen,Hehua Zhu,Haoxin Li,J. Woody Ju,Zhengwu Jiang,Zhiguo Yan
DOI: https://doi.org/10.1016/j.probengmech.2020.103064
IF: 2.954
2020-01-01
Probabilistic Engineering Mechanics
Abstract:The concrete properties inherently fluctuate even using the same manufacturing process. A Legendre polynomial based micromechanical framework is proposed to investigate the inherent randomness of the concrete's elastic behavior. The three-phase composite model is employed to represent the concrete material at microscale level. The volume fractions of different constituents are reached using a simplified analytical approach. Predictions with and without particle interactions for the concrete properties are obtained through the two-level homogenization process. By describing the microstructures with high-dimensional random vector, the stochastic micromechanical framework is obtained based on the presented three-phase model to incorporate the concrete's inherent randomness. The unbiased probability density functions (PDFs) of the properties are attained using efficient simulation framework consisting of the Legendre polynomial approximations, Monte Carlo simulations and the linear transformations. Numerical examples indicate that the proposed framework is able to describe the probabilistic characters of concrete properties from microscale level and obtain the material's unbiased PDFs with high efficiency. Finally, the statistical effects of ITZ are discussed on the concrete' s macroscopic probabilistic behaviors.
What problem does this paper attempt to address?