Fracture Behavior and Energy Analysis of 3D Concrete Mesostructure under Uniaxial Compression

Abstract

In order to investigate the fracture behavior of concrete mesostructure and reveal the inner failure mechanisms which are hard to obtain from experiments, we develop a 3D numerical model based on the Voronoi tessellation and cohesive elements. Specifically, the Voronoi tessellation is used to generate the aggregates, and the cohesive elements are applied to the interface transition zone (ITZ) and the potential fracture surfaces in the cement matrix. Meanwhile, the mechanical behavior of the fracture surfaces is described by a modified constitutive which considers the slips and friction between fracture surfaces. Through comparing with the experiments, the simulated results show that our model can accurately characterize the fracture pattern, fracture propagation path, and mechanical behaviors of concrete. In addition, we found that the friction on the loading surfaces has a significant effect on the fracture pattern and the strength of concrete. The specimens with low-friction loading surfaces are crushed into separate fragments whereas those with high-friction loading surfaces still remain relatively complete. Also, the strength of concrete decreases with the increase of the specimen height in the high friction-loading surfaces condition. Further, the energy analysis was applied to estimate the restraint impact of loading surfaces restraint on the compressive strength of concrete. It shows that the proportion of the friction work increases with the increase of the restraint degree of loading surfaces, which finally causes a higher compressive strength. Generally, based on the proposed model, we can characterize the complicated fracture behavior of concrete mesostructure, and estimate the inner fracture mode through extracting and analyzing the energies inside the cohesive elements.

The Effect of Inner Friction on Concrete Fracture Behavior under Biaxial Compression: A 3D Mesostructure Study

“…The Voronoi tessellation can generate seamlessly connected polyhedron cells based on the predefined seeds, and normally one seed can only generates one polyhedron, which provides a certainty for aggregate generation. In this paper, we adopt the aggregate-generating method which has been presented in our previous work [44]. The typical Voronoi polyhedrons and the generated aggregates are shown in Figure 1.…”

“…To characterize the mechanical behavior of ITZ and potential fracture surfaces (represented by cohesive elements), we adopted the modified constitutive model presented in our previous work with consideration of the friction effect between fracture surfaces [44], based on the mixed-mode damage model [49,50,51]. Here we give a brief introduction of this constitutive.…”

“…Besides the method mentioned above, a modified finite element method with cohesive elements is beginning to be used by more and more researchers [36,37,38,39,40,41,42,43,44]. The cohesive element is an interface transition element between solid elements.…”

“…Based on our previous research works about the numerical model with cohesive element [43,44], in this paper, we adopt the Voronoi tessellation and zero thickness cohesive element to generate concrete mesostructure, and a modified constitutive model considering the inner friction effect is also adopted. On the basis of this numerical method, we investigate the fracture behavior of concrete under biaxial compression in terms of the mechanical behavior, fracture pattern, and energy evolution.…”

Investigation of the Flow Properties of CBM Based on Stochastic Fracture Network Modeling

“…However, because of the complex structure of porous media and the limitations of considering real geological conditions, conventional experiments and theoretical achievements fail in describing the uncertainties of flow and transport properties in porous media. For example, the continuous equivalent model averages the permeability of reservoirs and overlooks the influence stemming from the discontinuity caused by fractures [11,12]. This model is similar to the “black box” and its central problem is to solve the permeability tensor.…”