Publication

o meet the increasing demands from regulatory bodies on CO2 emissions from automotive cars, the automotive industry is currently very active in reducing vehicle weight, where one significant step is to increase the amount of Fibre Reinforced Polymers (FRPs) due to their superior specific properties (e.g. specific strength and specific energy absorption in axial crushing) compared to conventional metals. However, as concluded in the European Roadmap to Safe Road Transport [1], the numerical finite element (FE) tools for the accurate prediction of the crash response of vehicle structures in FRPs are crucial for structural composites to have a widespread use in future cars. In order to achieve good predictability of the deformation of structural composite components in e.g. automotive crash and durability simulations, a proper modelling of the delamination process is crucial. Besides being a result of high interlaminar transverse (out-of-plane) stresses, delaminations are commonly initiated from stress concentrations at the tips of transverse matrix cracks, so-called matrix crack induced delaminations (MCID), illustrated in Figure 1. In fact, MCID is a common and challenging failure mechanism in laminated composites. In general, in order to capture delamination initiation and growth, detailed modelling of each ply by separate elements and interconnecting cohesive interface elements is required. However, due to restrictions on the simulation time, such high-fidelity models are not feasible to use in industrial applications such as full vehicle crash simulations. Therefore, in an attempt to achieve predictive and yet sufficiently efficient modelling of delamination initiation and growth, an adaptive enrichment methodology for the modelling of multiple and arbitrarily located delamination cracks using an equivalent single-layer (ESL) shell model has recently been presented [2]. The methodology has been shown to save substantial amounts of computational efforts, thus having the potential to enable computationally efficient simulations of progressive delamination failure in composite structures. However, up until recently, the methodology has not been able to capture the crucial mechanism of MCID.