From 3D X-ray CT scans to mesh-independent numerical compressive strength predictions of unidirectional carbon fiber composites

Pinelopi Mageira, Konstantinos Poulios, Jens W. Andreasen, Vedrana A. Dahl, Lars P. Mikkelsen

Abstract: This study presents a workflow to predict numerically the compressive strength of pultruded composites and the formation of kink bands, using material data acquired from mechanical testing and fiber orientation distribution obtained through the structure tensor analysis method on X-ray Computed Tomography. The mechanical compressive strength predictions were performed using GetFEM, a finite element framework for Python. Using that, a two-dimensional rectangular model is loaded in compression. The non-linear composite material is modelled using fiber-matrix homogenization. In here, the mesh independence is achieved through introducing the fiber’s diameter as a material’s length scale, capturing the fiber bending. The compression strength predictions are validated against composite testing. These predictions can be used for the adjustment of the pultrusion parameters for improved composite performance. In this work, a model case is presented along with the workflow for a more thorough understanding of the developed method.

External IDs:doi:10.1088/1757-899x/1338/1/012026