Investigation on progressive damage evolution for low-velocity impact simulation of woven composites

• Models with empirical and physical-based damage evolutions were assessed. • The work spans a wide energy range covering cases with and without penetration. • Mechanical response and damage morphology were investigated with experiments. • Pros and cons for different damage evolutions for composites were highlighted. This research aims at comparing the capability of three damage models, the enhanced composite damage model (MAT055), the Pinho laminated fracture model (MAT261), and the composite softening deformation gradient decomposition (DGD) model (MAT299) for woven composite materials, in predicting damage from low-velocity impacts. The first of them considers the empirical damage evolution with residual strength softening factors, and the other two control the damage evolution with fracture mechanism. To assess their predictive capabilities regarding mechanical response and damage, low-velocity impact (LVI) response of aramid-fibre epoxy plain-woven composites at four energy levels, from 27.9 J to 109.5 J, was investigated. A finite element model with macro-homogeneous solid element formulation was developed, and a rigorous calibration of the various physical and non-physical parameters was conducted (for all material models). Low-velocity impact tests were performed to identify the different failure mechanisms, focusing on the penetration of the impactor into the woven composites. The MAT261 with linear damage evolution better fits the experimental data at high impact energy levels, where it demonstrates high accuracy on mechanical response and damage propagation area. However, it requires significantly longer computational time. Overall, this study provides an in-depth understanding of the limitations and advantages of those material models, providing insight into their suitability to simulate the impact behaviour of woven composites.