Numerical prediction for viscoelasticity of woven carbon fiber reinforced polymers (CFRPs) during curing accounting for variation of yarn angle caused by preforming

Abstract

To model viscoelasticity of woven composites during curing, existing methods were mostly derived by directly mixing the material models for constituents and neglecting interaction among yarns, causing inaccuracy in numerical simulation. In addition, the preforming effects, which exist for production of parts with complex geometry, on curing of woven composites were rarely considered. In this paper, a novel geometric modeling method, involving finite element analysis (FEA) and element mapping, was first established to obtain voxel mesh for non-orthogonal representative volume element (RVE) structures. Through integration of the thermoviscoelastic constitutive models and voxel mesh, FEA was conducted to predict stress relaxation of woven CFRPs with varying yarn angles and degrees of curing (DOCs). Experimental validation indicates that the FEA can capture viscoelastic response of woven CFRPs with different yarn angles and DOCs with around 4.96% average weighed error, meaning the new approach can virtually characterize viscoelasticity of woven composites under complex processing conditions.