A multi-material-oriented modeling framework to characterize and predict mechanical self-healing

In the family of smart materials termed as “self-healing materials”, there is a prominent number of them sharing many similarities in the way that healing is affecting their post-damaged mechanical response. In this work, a generic and multi-material phenomenological-based healing formulation is proposed to investigate and characterize the self-healing effect in the mechanical response of materials exhibiting strong nonlinearities like rate-dependent plasticity, visco-damage initiation and evolution. The proposed healing formulation uses objective experimental measures such as resting time, loading rate and damage level. This formulation is integrated in a combined computational–experimental procedure, the so-called Generic Healing-Oriented Multi-Material Modeling Framework (GHOM ³) that facilitates (i) the understanding of a strongly nonlinear response, (ii) a robust material characterization and (iii) the predictive simulation via the finite element analysis. As study case, the intrinsic self-healable mechanical response of a highly nonlinear asphalt-based composite matrix is characterized at room temperature using the proposed framework. As additional novelty in the framework, an ultra-fast optimization-based material parameter identification process is developed using a master–slave parallelization paradigm that leads to save up to 90% of data processing time. The framework is put into practice to virtually predict the mechanical response influenced by the healing process of the validated material in a dog-bone specimen under Load–Unload–Resting–Reload at multiple loading rates and damage levels. The implemented approach gives direct access to the entire damage and healing histories, which are experimentally inaccessible, as well as providing an alternative definition of the healing indices commonly used in experimentation.