Open Access Open Access  Restricted Access Subscription or Fee Access

Automated Computed Tomography Segmentation of Impact- Damaged Curved Composite Structures



Carbon fiber composite structures are highly susceptible to impact damage which results in a complex 3D network of matrix cracking, delamination, and fiber failure. The difficulty of more conventional non-destructive evaluation techniques (e.g., ultrasonic pulse-echo scanning) in characterizing the full 3D damage state has prompted research interest in X-ray computed tomography (CT), which can provide a 3D map of the internal structure and damage. However, the large size of CT datasets and the complexity of damage features can complicate the interpretation of damage as being structurally relevant damage modes or not. Additionally, geometric details such as acute curvature can complicate damage interpretation. This paper presents a CT segmentation procedure that enables the segmentation of datasets of curved composite structures. This procedure will be demonstrated on stringer stiffened composite panels with the layup [45/-45/0/45/90/-45/0/90]s impacted on the stringer cap with a 25 mm radius impactor. The segmentation procedure outputs 2D maps of the damage for each ply and interface and outputs 3D datasets of color coded damage overlaid onto the original CT dataset. Results from the segmentation can also be combined into 3D visualizations of the damage state.


Full Text:



  • There are currently no refbacks.