PSI - Issue 77

Kumar C. Jois et al. / Procedia Structural Integrity 77 (2026) 405–412 Jois, et al./ Structural Integrity Procedia 00 (2026) 000–000

407

3

distribution across the laminate thickness—information that is difficult to obtain through traditional microscopy or destructive sectioning.

(a)

(b)

Hough Lines

Y

20 mm

Z

Fig. 1. (a) Image of voids extracted from GeoDict; (b) Hough line generation through image processing in Python.

For the analysis of voids and their spatial distribution within the composites, 3D scan data obtained from computed tomography were post-processed to extract the void size and location. These data were then projected onto a 2D plane using a Python-based image processing library, see Fig. 1. Depending on the selected plane, the orientation of voids was determined using the Hough line transform method, which identifies linear features within the image. This approach enables quantitative evaluation of both the void density within each layer and the preferred orientation of voids, providing valuable insight into the relationship between manufacturing parameters and defect distribution. 2.2. Numerical model and boundary condition To analyse the influence of voids on the structural behaviour of the composite pressure vessel, finite element models were developed. To accurately represent the full laminate configuration without introducing cohesive interfaces, appropriate loads and boundary conditions were defined, as illustrated in Fig. 2. A liner was included in the model and acted as the surface to which an internal pressure load of 200 MPa was applied, representing the operational burst pressure.