Three-dimensional X-ray tomography for multiscale characterization and fatigue damage investigations of composites for wind turbine blades.

Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

32 Downloads (Pure)


Key mechanical properties of the load carrying composite laminates in wind turbine blades are the stiffness, fatigue resistance and the compression strength. While the overall design criteria for glass fiber composites mainly are stiffness driven, it tends to be more fatigue and compression strength driven when using carbon fiber or glass/carbon hybrids. Using lab-source x-ray computer tomography in connection with ex-situ fatigue testing of quasi uni-directional composites, local fiber failure has been followed. The material investigated is typical non-crimp fabric used in wind turbine blades. The fatigue damage evolutions is studied during tension/tension, tension/compression and compression/compression fatigue. It is found that the location of the tension/tension (R=0.1) and tension/compression (R=-1) fatigue damage are controlled by the bundle architecture, while the compression/compression (R=10) show up as small shear cracks located in the matrix between the fibers. In addition to fatigue damage investigations, the fiber architecture inside the fiber bundle and the fiber bundle architecture is studied. Based on a dictionary based segmentation method, the fiber-architecture inside the fiber bundle is explored. The fiber bundle architecture is studied using more conventional segmentation methods. It is segmentations, which can used as input to multi-scale finite element predictions, e.g. aiming for a predictions of the stiffness and the compression strength reductions due to fiber misalignment and local stress distribution causing fatigue damage due to the architecture of the fiber reinforcement. Three-dimensional X-ray computer tomography large amount of data, typical between 2-50GB from each dataset. During publication, only a small amount of such data set can be shown. Therefore conclusions from such studies can be doubtful, if not the full data-set is made available together with the publication. Strategies for how to do this will be presented at the end of the talk.
Original languageEnglish
Publication date2018
Publication statusPublished - 2018
EventCivil Engineering Seminar Series (CESS) - Lausanne, Switzerland
Duration: 21 Dec 201821 Dec 2018


SeminarCivil Engineering Seminar Series (CESS)

Cite this