The susceptibility of structural concrete to early-age cracking depends on material composition, methods of processing, structural boundary conditions, and a variety of environmental factors. Computational modeling offers a means for identifying primary factors and strategies for reducing cracking potential. Herein, lattice models are shown to be adept at simulating the thermal-hygral-mechanical phenomena that influence early-age cracking. In particular, this paper presents a lattice-based approach that utilizes a model of cementitious materials hydration to control the development of concrete properties, including stiffness, strength, and creep resistance. The approach is validated and used to simulate early-age cracking in concrete bridge decks. Structural configuration plays a key role in determining the magnitude and distribution of stresses caused by volume instabilities of the concrete material. Under restrained conditions, both thermal and hygral effects are found to be primary contributors to cracking potential.

Short fiber reinforcement is added to concrete materials to improve a variety of performance measures related to structural safety, serviceability, and health diagnosis. Mesoscale models have been developed to understand the individual and collective actions of these fibers on various material properties. Those modeling efforts have predominately focussed on the mechanical (i.e., stiffness and strength) contributions of fibers. This paper introduces computationally efficient, semi-discrete representations of fibers within coupled mechanical and transport processes in cement-based matrices. Basic simulations are done to study: the use of conductive fibers for self-sensing; and the influences of fibers on early-age plastic settlement. It is found that the models can account for directional bias on fiber orientation, as may occur during material casting. With respect to plastic settlement, fibers may play competing roles: mechanical restraint offered by the fibers reduces settlement, whereas enhanced hydraulic conductivity along the fiber–matrix interface may increase settlement by facilitating the bleeding process.

This research involves the multiscale characterization of strain-hardening cementitious composites under tensile loading. The sensitivity of cracking behavior to fiber dispersion is studied using three-dimensional lattice models, in which each fiber is explicitly represented. It is shown that the nonlocal modeling of fiber bridging forces is essential for obtaining realistic patterns of crack development. Crack count and crack size are simulated for progressively larger levels of tensile strain. The influence of fiber dispersion is clearly evident: regions with significantly fewer fibers act as defects, reducing strength and strain capacity of the material.