Extrinsic self-healing materials are materials with built-in (micro-) capsules or vessels, which upon fracturing release healing agents in order to recover the material's physical and mechanical properties. In order to better understand and engineer these materials, a thorough characterization of the materials' microstructural behaviour is essential and often overlooked. In this context, we illustrate how micro-computed tomography (µCT) can be used to investigate the distribution and debonding of (micro-) capsules in their native state and in three dimensions in a polymer system with self-healing properties. Furthermore, we show how in-situ µCT experiments in a self-healing polymer and a selfhealing concrete system can elucidate the breakage and leakage behaviour of (micro-) capsules at the micrometre scale. While challenges related to image resolution and contrast complicate the characterization in specific cases, non-destructive 3D imaging with µCT contributes to the understanding of the link between the microstructure and the self-healing behaviour of these complex materials.

Long-term predictions of material properties such as stiffness and creep resistance are important in many engineering applications and require high reliability and accuracy. This is especially true for polymer materials and their composites as their viscoelastic nature results in timedependent material behaviour and any measurement uncertainties or errors amplify in longterm predictions. To measure this behaviour at smallest loadings, Dynamic Mechanical Analysis (DMA) is frequently declared as an ideal method. However, the measurement accuracy and repeatability of this method is strongly influenced by (i) the testing fixture and corresponding loading mode, (ii) the sample preparation and (iii) the plotting scale to interpret the test results. In this study, relevant experimental parameters were found for DMA and a proper procedure was designed, which was then applied to measure the viscoelastic behaviour of a highly temperature and creep resistant thermoplastic polymer (polyethersulfone) and of a highly graphite filled polypropylene composite. In combination with finite element simulations and in-situ strain measurements by digital image correlation (DIC), the main influences on measurement accuracy of three-point-bending DMA were identified and subsequently used to determine measurement guidelines. Using these guidelines, DMA measurements allow

No abstract