A review on fracture propagation in concrete: Models, methods, and benchmark tests

Rigorous characterization of fracture evolution in quasi-brittle materials – such as concrete – requires computationally intensive and, mostly, complex approaches. However, the headway in computational power and the efficiency of numerical methods increased the interest in the numerical modeling of concrete fracture. This paper, first, reviews concrete fracture from different perspectives of fracture theories/models based on macroscopic or mesoscopic formulations along with thorough discussions on the existing numerical methods for their solutions. Next, a diverse set of both classical and newly developed experimental concrete fracture tests is compiled and systematically critiqued along with past and evolving literature numerical simulations of the same. This serves as a state-of-the-art benchmark suitable for validating newly emerging numerical models and/or fracture theories. The work, then, discusses the size effect laws controlling the nominal strength of concrete structural elements and analyzed existing studies on the influence of mixture design on the concrete fracture response. Based on the conclusions of the review on numerical models, fracture theories, the compiled experimental benchmarks/tests, and the factors controlling concrete fracture, the literature gap is established and future perspectives highlighted. Finally, the paper develops a series of numerical models based on 3-D adaptive generalized finite element method (GFEM) to validate its ability in simulating mixed-mode concrete fracture trajectories of various benchmarked problems. The GFEM's superiority in terms of mesh adaptivity, ability to use coarse and unstructured meshes (contrary to the use of expensive discretizations by most of the existing models), as well as resulting in a stable system of the generated equations are established.