A continuous-discontinuous strategy to represent the crack process in concrete structures

Resumo

The study of fracture in quasi-brittle materials such as concrete has significant importance since it is
one of the main causes of material failure. There are two numerical approaches to representing fracture: smeared
and discrete models, and both techniques have pros and cons. Among the smeared approaches, damage models
are used to reproduce the degradation of a continuum media. These models are appropriate for describing the first
stages of degradation, identifying damaged regions, and replacing original mechanical properties with damaged

ones. This strategy assumes that the cracks are spread over an area known as the fracture process zone. How-
ever, this phenomenological approach cannot represent the crack path properly since the discontinuities are not

geometrically described. In contrast, the discrete methods are the most indicated to characterize the fracture ex-
plicitly. Such methods frequently deal with remeshing, an alternative that has been avoided because of the high

computational cost. Although, based on the efficiency of modern computers, it is now possible to evaluate the
viability of coupling continuous and discontinuous models to reproduce fracture in its entirety, from nucleation to
collapse. In this context, it is proposed a combined strategy that associates nonlocal damage models to represent
the smeared aspects of crack propagation with a discrete crack description based on nodal duplication to capture
the crack discontinuity. Finally, numerical simulations were performed to analyze the efficiency of this strategy in
representing the degradation processes from smeared cracks to geometric discontinuities.