Abstract
The mechanical properties of concrete composites mainly depend on its ingredients and meso-structure, especially the shape, distribution, and grading of aggregates. Therefore, it is very important to establish realistic meso-scale models for concrete composites performance prediction and structural optimization design. However, there is only a few methods to generate models with high volume fraction in current studies. In this paper, a novel idea for numerical modeling, namely the aggregate expansion method, is used to establish more realistic 3D models with different volume fractions of convex and concave aggregates. The shrunken aggregates are first placed in an empty container, and then these aggregates are expanded to the predetermined sizes to obtain a numerical model. Using this approach can efficiently obtain a high volume fraction model with randomly distributed aggregates. In addition, realistic aggregates are generated according to the statistical data of actual aggregates shapes. A series of statistical data show that the randomness of aggregates in the space before and after expansion is still maintained. At last, the simulation results of 3D numerical models with different volume fractions under uniaxial tension are calculated. The consistency between the simulation results and experimental data verifies the good accuracy of the models.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call