RESUMO
We performed in-situ tensile tests on two carbon fibre/epoxy composites with continuous scanning using synchrotron computed tomography (CT). Both composites were cross-ply laminates, and two specimens were tested for each composite. The voxel size was sufficiently small to recognize individual fibres and fibre breaks. For each test, 16-19 volumes were reconstructed, cropped down to the 0° plies and analysed to track fibre break and cluster development. This dataset provides the last CT volume before failure for each of the four specimens as well as the individual fibre break locations in all reconstructed volumes. These data are then plotted against predictions from six state-of-the-art strength models. The target is that these data become a benchmark for the development of new models, inspiring researchers to set up refined experiments and develop improved models.
RESUMO
X-ray microfocus computed tomography (micro-CT) is recently put forward to qualitatively and quantitatively characterize the internal structure of porous materials. However, it is known that artifacts such as the partial volume effect are inherently present in micro-CT images, thus resulting in a visualization error with respect to reality. This study proposes a validation protocol that in the future can be used to quantify this error for porous structures in general by matching micro-CT tomograms to microscopic sections. One of the innovations of the protocol is the opportunity to reconstruct an interpolated micro-CT image under the same angle as the physical cutting angle of the microscopic sections. Also, a novel thresholding method is developed based on matching micro-CT and microscopic images. In this study, titanium porous structures are assessed as proof of principle. It is concluded for these structures that micro-CT visualizes 89% of the total amount of voxels (solid and pore) correctly. However, 8% represents an overestimation of the real structure and 3% are real structural features not visualized by micro-CT. When exclusively focusing on the solid fraction in both the micro-CT and microscopic images, only an overestimation of about 5% is found.
