Compressive failure predictions for carbon fiber-reinforced composites: A comparative study between imaging modalities

We compare compressive strength predictions for carbon fiber-reinforced composites based on different imaging modalities. We use X-ray computed tomography and optical microscopy for acquiring 3D and 2D images, respectively. Based on the acquired images, the fiber orientation distribution is estimated and used as input for 2D and 3D finite element models of both imaging modalities. Subsequently, the four model predictions are compared with experimental data to quantify their capabilities in terms of stiffness and strength. All models predict compressive failure associated with kink band formation near areas with a high degree of fiber misalignment. The 3D models predict a stiffness within one standard deviation of the experimental results but with significantly higher strength. We argue that the strength prediction shows a potential, as the measured strength is limited by load introduction failure. The 2D models predict a stiffness 4-7% higher than the 3D models. The strength predictions are similar to the measured strengths but 32-48% lower than the prediction based on 3D X-ray μCT. The discrepancy between strength predictions shows that a simple 2D approach is too limited for the investigated composite and that 3D models based on more advanced imaging modalities are required to accurately predict compressive failure.