Abstract
1- Introduction
2- Generation of 3D mesoscale models for plain and fiber reinforced concretes
3- Mesoscale models of plain and fiber reinforced concretes
4- Properties analysis on 3D mesoscale models for plain and fiber reinforced concretes
5- Conclusions
Acknowledgment
References
Abstract
Model construction is a key and indispensable step to understand the internal structure of concrete at the micro-/mesoscale level, which affects its properties in practice. In this paper, a practical and applicable method is proposed for generating the mesoscale structure of plain and fiber reinforced concretes. In this method, cell fracture algorithm was developed to obtain arbitrary-shaped aggregates, Surface subdivision (Catmull–Clark subdivision algorithm), Displacement mapping and Laplace smoothing algorithm were developed to constructed rough surface of realistic aggregates. Random algorithm was used to generate fibers, the interactions between aggregates and fibers were detected and solved by collision algorithm. The influence of shape, size and volume fractions of aggregate, together with fiber’s orientation on the structure and properties of plain and fiber reinforced concrete were studied. Compared with experimental data and previous works, the proposed models can well predict the volume fraction of the interfacial transition zone (ITZ) and the elastic modulus of plain and fiber reinforced concretes. This paper provides a promising tool for numerical and analytical research of plain and fiber reinforced concretes.
Introduction
Concrete is a multi-phase and heterogeneous material. It consists of three important components: cement paste, aggregate and the interfacial transition zone (ITZ). Fiber is often added into concrete in practice to improve its mechanical performance, including the flexible strength, toughness, etc. [1]. At mesoscale, fiber-reinforced concrete (FRC) can be simplified as aggregates and fibers embedded in the cement matrix, whereas the ITZ is the interface between different phases. Still, it is very difficult to precisely understand the realistic structure of plain and fiber-reinforced concretes by traditional experimental methods without destroying their structures. X-ray tomography image-based reconstruction technique [2, 3] is a promising approach to solve this problem, however, this new technique requires specialized instruments and is time-consuming, laborious and thus expensive. Moreover, the resolution of this approach has an important effect on the accuracy of the obtained results, which limits its application in cementitious materials. As an alternative, numerical modeling provides an economic and reliable way to construct the concrete structure at mesoscale.