Non-linear FE2 multiscale simulation of damage, micro and macroscopic strains in polyamide 66-woven composite structures: Analysis and experimental validation

Abstract

This paper presents an experimental approach aimed at analyzing and validating a two-scale nonlinear Finite Element (FE 2 ) simulation of a 3D composite structure. The studied composite material consists of polyamide thermoplastic matrix, exhibiting viscoelastic-viscoplastic behavior with ductile damage, reinforced by woven glass fabric whereby inelastic and anisotropic damage behavior is considered. The multiscale parallel computation is founded on the periodic homogenization at the microscopic scale, which considers the geometric description of the reinforcement's architecture and accounts for time-dependent and non-linear local behavior of

Hybrid hierarchical homogenization theory for unidirectional CNTs-coated fuzzy fiber composites undergoing inelastic deformations

“…An additional simplification employed in the present work is that the inelastic behavior of the NCP is described directly using the plastic flow rule for transversely isotropic material. Parametric investigations and applications explicitly accounting for the nonlinear deformation mechanisms of the CNTs and the surrounding matrix require a multiscale analysis methodology such as FE 2 analysis [35], which will be addressed in a forthcoming publication.…”

“…The finite-element [28][29][30] and finite-volume [31][32][33] methods in their various forms are the dominant techniques because of the ability to admit complex unit cell structures containing random inclusions of arbitrary shapes and sizes. They also permit the incorporation of arbitrary deformation mechanisms at the individual phase, such as plasticity [34], viscoelasticity-viscoplasticity [35,36], martensitic transformation [37], as well as hyperelasticity in finite transformation [38]. For the case of a continuous fuzzy fiber-reinforced composite, the effective properties have been determined by Kundalwal and Ray [19,39] using the method of cells approach and the finite element method.…”

Viscoelastic-viscoplastic homogenization of short glass-fiber reinforced polyamide composites (PA66/GF) with progressive interphase and matrix damage: New developments and experimental validation

“…These models, in general, can be categorized into two broad categories. The first class of approaches is based on the full-field analysis, such as the finite-element (Chatzigeorgiou et al, 2016;Praud et al, 2021;Tikarrouchine et al, 2021), finite-J o u r n a l P r e -p r o o f 7 volume (Chen et al, 2017;Chen, 2020), asymptotic (Cruz-González et al, 2020;Rodríguez-Ramos et al, 2012), locally-exact (Drago and Pindera, 2008;He and Pindera, 2021), mesh-free (Chen and Aliabadi, 2019), and fast Fourier transformation (Lahellec and Suquet, 2007;Moulinec and Suquet, 1998) homogenization of composite materials containing elastic, viscoelastic, viscoplastic, and damage phases. The full-field analysis approaches are suitable for modelling detailed microstructures but require substantial effort for the input data construction.…”

An efficient monolithic solution scheme for FE$^2$ problems

“…While the implementation of the sequential scheme is comparatively simple, the concurrent approach is challenging and computationally expensive, but still quite universally applied. To name a few examples, it has been used in numerical multi-scale analysis of fibre-reinforced composites [9,29,36], woven composites [35,38], elastic-plastic matrix-inclusion problems [20,29], or the growth of microvoids [16,20] and texture formation in polycrystalline metals [22], and in the design optimization of microstructures [24]. Following much research attention over the last two decades, the FE 2 method is no longer restricted to conventional mechanical problems.…”

“…Kouznetsova et al [16] presented an implementation in 2001, where the macroscopic model is computed by a MATLAB FE-code and the microscopic problems are solved by the commercial package MARC [16]. Yuan and Fish [38], Tchalla et al [33] and Tikarrouchine et al [35,36] utilize Abaqus for both scales with the aid of a python script and a procedure UMAT to let the Abaqus macro program calls itself at each integration point. Naturally this listing is not complete.…”