An efficient geometric reconstruction of mesoscale concrete structures accounting for confinement scenarios

An efficient numerical modeling of concrete composites requires both the definition of the correct geometrical reconstruction of samples and the development of a sound model for the service and ultimate loads. This work proposes a procedure (characterized by the adoption of laser scanner techniques combined with CAD, X-ray computed tomography, and a particle random disposition algorithm) for reproducing a finite element model of concrete at the mesoscale in terms of particle packing and representation of material heterogeneities, together with a nonlinear constitutive characterization developed in the framework of nonassociated plasticity coupled with damage. The approach is validated against a typical experiment of structural collapse of concrete samples subjected to uniaxial compression, where the cement paste fails under aggregates confinement.