Simulations of delamination in CFRP laminates: Effect of microstructural randomness

Abstract

Due to their high specific strength and stiffness, fibre-reinforced composite materials are being increasingly used in structural applications where a high level of performance is important (e.g. aerospace, automotive, offshore structures, etc.). Performance in service of these composites is affected by multi-mechanism damage evolution under loading and environmental conditions. For instance, carbon fibre-reinforced laminates demonstrate a wide spectrum of failure mechanisms such as matrix cracking and delamination. These damage mechanisms can result in significant deterioration of the residual stiffness and load-bearing capacity of composite components and should be thoroughly investigated.The delamination failure mechanism is studied in this paper for a double cantilever beam (DCB) loaded in Mode I. Several sensitivity studies are performed to analyse the effects of mesh density and of parameters of the cohesive law on the character of damage propagation in laminates. The microstructural randomness of laminates that is responsible for non-uniform distributions of stresses in them even under uniform loading conditions is accounted for in the model. The random properties are introduced with the use of Weibull's two-parameter probability density function. Several statistical realisations are carried out which show that the effect of microstructure could significantly affect the macroscopic response emphasizing the need to account for 2 microstructural randomness for accurate predictions of load-carrying capacity of laminate composite structures.

Three-Dimensional Analysis of the Effect of Material Randomness on the Damage Behaviour of CFRP Laminates with Stochastic Cohesive-Zone Elements

“…More detailed analyses were performed based on three-dimensional models and a number of statistical realizations based on a half-scatter of 50% of fracture energy were presented. In contrast to the two-dimensional analyses reported previously [8], the results here showed lower load-bearing capacities for most of the random models than that in the uniform model. Also, the damaged area and the crack lengths also showed higher values in the random realizations compared to the uniform case above a certain value of applied displacement.…”

“…A considerable difference in the values of reduction in stiffness between cases with different crack spacings suggested that the assumption of averaged distributions of defects can lead to unreliable predictions of structural response. In another work by Khokhar et al [8], based on twodimensional DCB models, it was concluded that the variation in the fracture properties undertaken to analyze the effect of material randomness using stochastic cohesive zone elements had a potentially significant effect on the predictions of the performance of CFRP laminates.…”

“…In addition, the previous literature in this area indicates the need for a more detailed analysis based on three-dimensional simulations. In the previous two-dimensional analysis [8], the variation of fracture parameters within the cohesive layer was only one-dimensional, i.e., only along the length of the double-cantilever beam. To have a more complete picture of the effect of material randomness on the specimen under study, a three-dimensional analysis is necessary in order to investigate the effect of twodimensional variations in the microstructure in the plane of delamination.…”

Finite-element modelling of bending of CFRP laminates: Multiple delaminations

“…Thus, the application of CZE requires a fine spatial discretization at the cohesive zone to capture the damage growth properly. However, such refinement may be prohibitive since it needs a significant increase in computational efforts [26]. An optimum number of elements are required in the cohesive zone to obtain accurate numerical results.…”

Development of Delamination in Cross-Ply Laminates: Effect of Microstructure

“…Case C. Model with one crack and cohesive elements between 0 o and 90 o layers using random fracture properties based on a 2-parameter Weibull's distribution [8] (multiple realizations presented). A value of half scatter of 50% of properties is selected to underline the effect of the material's stochastic nature on the course of damage initiation and propagation.…”