Gradient enhanced multi-scale modeling framework for glass fiber reinforced polyamides

Abstract

This study proposes a multi-scale gradient enhanced nonlocal modeling framework aimed at predicting the mechanical response of long glass fiber reinforced polyamide composites that exhibit nonlinear viscoelastic viscoplastic rheology with ductile damage driven by plasticity. To treat numerically material instabilities at severe damage levels, an internal length scale is introduced within the model through a gradient enhanced framework that controls the non-physical localization of state variables and the consequent early model failures. To do so, a viscoelastic viscoplastic phenomenological model is adopted to capture the matrix phase nonlinear response at the microscale, then an appropriate homogenization approach is adopted to provide the overall response of the composite, in which the gradient enhanced framework is imposed at the macroscale. As a result, a consistent homogenization model capable of capturing nonlocal phenomena is introduced and implemented in a commercial finite element software, which addresses the non-physical responses of the local model and exhibits higher stability.