X-ray holotomography unveils microscale damage in polymer composites

Fibre-reinforced polymer composites (FRPCs) offer outstanding mechanical properties and low density, making them essential in aerospace and energy sectors and increasingly important in automotive applications. Their mechanical performance depends on the effective transfer of stress between fibres and the matrix, which is governed by the fibre-matrix interface.

Interfacial strength is a critical parameter often evaluated using micromechanical test methods [1]. These tests typically employ single-fibre composites, where stress is applied to the interface between a single fibre and the surrounding matrix. However, such methods assume that damage mechanisms in single-fibre configurations reflects those in multi-fibre composites, which feature higher fibre volume fractions in unidirectional plies. This assumption, however, has not been experimentally validated.

The heterogeneous microstructure of composites make them highly suitable for X-ray computed tomography. Yet, high-resolution studies of the fibre-matrix interface remain challenging due to the required spatial resolution. In this study, in-situ tensile tests were conducted at the ID16B beamline, allowing for precise detection and monitoring of interfacial damage. The beamline enables in-situ holotomography, a phase-contrast X-ray nano-imaging technique capable of 3D and 4D imaging with a voxel size as small as 25 nm.

Single-fibre and multi-fibre specimens were prepared using either glass or carbon fibres embedded in an epoxy matrix doped with nanoparticles to enable strain mapping via digital volume correlation. Detailed descriptions of the manufacturing process and assessments of the mechanical and rheological properties of the nanoparticle-doped epoxy can be found in [2].

Tensile loading was applied using a custom-designed 1kN rig, and holotomography scans were performed at specific stress levels, as illustrated in Figure 1. The reconstructed tomograms were analysed to extract critical damage parameters, including apparent critical length, debond and interdebond length, matrix crack volume and maximum propagation, and the number and spacing between fibre breaks.
 

Fig. 1: Main components of the holotomography setup used at beamline ID16B for in-situ tensile tests on single- and multi-fibre composites.

Three primary damage mechanisms were identified and analysed:

1. Fibre fracture: Single-fibre specimens exhibited one to two fibre breaks within the field of view. Glass fibre breaks formed inclined planes relative to the loading axis (Figure 2a), whereas carbon fibre breaks were planar and perpendicular to the loading direction. Multi-fibre specimens displayed multiple fibre breaks, with fracture morphologies resembling those seen in single-fibre specimens. Variations between glass and carbon fibres were attributed to their structural differences.

2. Matrix cracking: Extensive matrix cracking was observed in glass single-fibre specimens, propagating in three distinct directions. The primary crack extended from the converging side of fibre fracture cone, while two smaller crack fronts emanated from the diverging side (Figure 2a). Conversely, glass multi-fibre specimens exhibited negligible matrix cracking (Figure 2b) despite multiple fibre breaks, with crack volumes four times smaller under comparable stress levels. No matrix cracking was detected in carbon fibre specimens in either single- or multiple-fibre configurations.

3. Interfacial debonding: Significant interfacial debonding was observed in single-fibre specimens for both fibre types, characterized by axisymmetric but non-uniform patterns along the fibre circumference (Figure 2a). In multi-fibre specimens, no interfacial debonding was detected (Figure 2b).

Click figure to enlarge

Fig. 2: Differences in damage mechanisms between (a) glass single-fibre specimens and (b) glass multi-fibre specimens. UTS: Ultimate Tensile Strength.

When fibre fractures occur in polymer matrices, the released energy is dissipated through matrix cracking and interfacial debonding. The prevalence of each mechanism depends on the mechanical properties of the fibre-matrix interface. Strong interfaces favour matrix cracking, while weaker interfaces promote debonding. The absence of matrix cracking and interfacial debonding in multi-fibre composites indicates fundamental mechanistic differences that require further investigation.

The observed differences can be attributed to four key factors. First, the smaller matrix volumes in multi-fibre specimens reduce defect volumes, increasing resistance to damage. Second, neighbouring fibres influence polymer curing and molecular arrangement, thereby enhancing interfacial properties. Third, multi-fibre configurations create a stiffer environment, which restricts crack propagation and energy dissipation. Finally, variations in local strain states between single- and multi-fibre configurations affect strain energy release and the associated damage mechanisms.

In conclusion, this study questions the reliability of single-fibre composites for accurately characterizing interfacial properties. The complex interactions present in multi-fibre specimens, which are absent in single-fibre configurations, highlight the need for more representative testing methods to evaluate interfacial properties in fibre-reinforced polymer composites.

Principal publication and authors
Questioning the Representativeness of Damage Mechanisms in Single-Fiber Composites via In Situ Synchrotron X-Ray Holo-Tomography, T. Chatziathanasiou (a), Y. Lee (b) O. Stamati (c), J. Villanova (c), S. AhmadvashAghbash (a), B. Fazlali (a), C. Breite (a), I. Sinclair (b), M.N. Mavrogordato (b), S.M. Spearing (b), M. Mehdikhani (a), Y. Swolfs (a), Small 2406168 (2024); https://doi.org/10.1002/smll.202406168
(a) Department of Materials Engineering, KU Leuven, Leuven (Belgium)
(b) μ-VIS X-ray Imaging Centre, University of Southampton, Southampton (UK)
(c) ESRF

References
[1] S. AhmadvashAghbash et al., Int. Mater. Rev. 68(8), 1245-319 (2023).
[2] T. Chatziathanasiou et al., Comp. B: Eng. 276 (2024).