| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1561155 | Computational Materials Science | 2013 | 14 Pages |
This study aims at generating numerical 3D samples of concrete so as to study the effects of the granular inclusions shape on the macroscopic kinetics of reactive transport phenomena. Two types of mesostructure configurations are considered: the first one is composed of a matrix of mortar in which are randomly distributed inclusions corresponding to the concrete coarse aggregates, and the second one also includes a steel rebar. The choice of a mesoscopic modeling for the mortar matrix is based on the need to obtain numerical structures of reasonable size. In particular, the Interfacial Transition Zones (ITZs) are assumed to be incorporated into the homogenized mortar properties. This study is applied to the case of drying and atmospheric carbonation by using simplified models solved by the finite element code Cast3M. The purpose is to quantify the influence of the aggregate shape on the kinetics of macroscopic transfer and the isovalue lines for some physical variables representative of the reactive transport problems: saturation degree for drying, and porosity, calcite and portlandite concentrations for carbonation. Basic aggregates shapes are studied (spheres, cubes), as well as more complex ones (Voronoi particles) which are supposed to be more representative of real aggregates. The effects of ‘non-isotropic’ shapes (oblate and prolate ones) are also investigated. It is shown that the influence of the aggregate shapes appears negligibly small on macroscopic indicators, except for oblate shapes with aspect ratios of 3. This latter case also exhibits substantial local delayed effects and a more important variability, which may have some importance for a precise description and estimation of degradation processes related to steel rebar corrosion.
► 3D numerical samples of concrete with aggregates of various shape are generated. ► FE simulations of drying and carbonation are carried out with simplified models. ► Isotropic aggregate shapes have negligible effects on macroscopic indicators. ► Oblate shapes with aspect ratios of 3 affect significantly the results. ► Shapes should be considered when precise state variable values are needed locally.
