P.H. Rios Silveira
Please Note
1 records found
1
Micromechanical models offer a physics-based alternative to phenomenological approaches to simulate the behavior of quasi-brittle materials. By combining mean-field homogenization techniques with fracture mechanics principles, these models aim to capture anisotropic damage evolution, unilateral effects, and multiphysics coupling with a small number of inputs. However, despite their theoretical appeal, their practical application is often hindered by critical limitations. This paper presents a comprehensive and critical examination of micromechanical formulations, focusing on the influence of homogenization schemes, damage evolution criteria, and loading type in model response. It highlights key issues such as the limited accuracy of homogenization estimates at high crack densities, the instability of post-peak responses, spurious damage localization, and the challenges of modeling tensile-compressive asymmetry and non-homothetic crack growth. Through analytical derivations and numerical examples, the study demonstrates that many micromechanical models rely on assumptions that break down under more general conditions, leading to non-physical predictions outside the scope of model conception. The findings suggest that while micromechanical models are valuable in specific contexts, their broader applicability requires careful scrutiny and further development.