Analysis of laminar fiber composites with computed tomography using shape-fitted layers

Fiber composites often consist of several layers with different fiber orientations and are formed in a three-dimensional shape. For non-destructive testing of these components industrial computed tomography (CT) can be used to analyze the internal structures and to detect inhomogeneities or defects. A set of several hundred X-ray projection images is acquired from which the local attenuation coefficients of the components are reconstructed to a three-dimensional volume. Difficulties arise when the three-dimensional volume representation of a laminar component is visualized. Often these components are curved and thus it is difficult to analyze the individual layers of the laminate by scrolling through the slices of the Cartesian volume. This presentation describes a method for the extraction of laminate layers from CT data and their subsequent evaluation. The surface of the component is described by Bézier-patches. Shifting and sampling of this surface description allows the calculation of a virtually "planar" volume. The laminate layers are aligned in parallel planes in the resulting volume and can be analyzed and visualized layer by layer. With the help of subsequent image analysis applied to each layer irregularities in the fiber layout like undulations or defects like pores and voids can be detected. The application of this method is demonstrated on a typical example.