Hidden flaws in carbon-fibre composites: how tiny defects decide outcomes

Carbon-fibre-reinforced polymer composites are integral to modern engineering, supporting structures such as airliner wings, hydrogen tank vessels, racing car monocoques, and wind turbine blades, all while maintaining a lighter weight compared to steel or aluminum. However, a recent review in the Journal of Materials Science highlights that the reliability of these composites is compromised by microscopic and mesoscopic defects inherent in the manufacturing process. These defects, and their interactions, can evolve into damage that affects the component's performance over its service life.

The review, conducted by Ziyang Liu and Minghe Li from Jilin Agricultural Science and Technology College, along with their team, explores the formation of process-induced defects, their quantitative characterization, and their role in initiating and propagating damage. The authors propose viewing defects not as isolated flaws but as local structural perturbation fields shaped by manufacturing processes. The impact of these defects is determined by their influence on local fibre architecture, interlaminar constraints, and stress transfer pathways.

Voids are the most common perturbation fields, forming during resin impregnation, curing, and consolidation when air, volatiles, or moisture become trapped in the polymer matrix. X-ray computed tomography has advanced the field by enabling three-dimensional mapping of voids, revealing their spatial morphology and clustering behavior. Studies cited in the review indicate that voids can independently cause matrix cracking, interfacial debonding, and interlaminar delamination, with their effects heavily influenced by their position relative to ply interfaces and stress concentrations.

Fibre waviness presents another challenge. Misaligned fibres, due to layup errors or cure-induced deformation, can significantly reduce the composite's compressive strength. The review discusses how embedded wrinkles can lead to kink-band formation under compression and how waviness concentrates damage, reducing fatigue life. The interaction between fibre waviness and voids further accelerates failure, as shown in recent analyses of unidirectional CFRP under compression.

Automated fibre placement (AFP), a key process in aerospace composite manufacturing, introduces defects such as tow-to-tow gaps and steering-induced wrinkles. The review references experimental studies showing that AFP defects can reduce shear and compression properties and alter damage development. These defects are distributed over large areas, making their cumulative effect challenging to assess with traditional single-defect acceptance thresholds.

A significant finding of the review is the non-additive nature of multiple defects. Two defects that individually cause minor property reductions can collectively lead to a significant loss of strength or partially compensate for each other, depending on their spatial arrangement. This non-additivity suggests that conventional quality control methods are inadequate for assessing the reliability of complex composite structures. The authors advocate for a probabilistic multidefect and multiscale framework that considers the stochastic distribution of defects and their interactions.

Emerging tools such as computational micromechanics and stochastic multiscale analysis could support this framework. Explicit defect modeling allows for the integration of realistic defect geometries into finite element models, tracking damage initiation and propagation. Stochastic models can propagate microvoid randomness to laminate-level behavior, facilitating uncertainty quantification for large structures. Data-driven methods, including deep-learning segmentation and machine-learning identification, are enhancing the analysis of experimental data.

The review suggests a shift from post-manufacturing screening to proactive defect control during formation. Process monitoring, digital twins, and online repair frameworks are identified as key components of a closed-loop manufacturing paradigm. Standardized defect characterization and data-reporting frameworks, aligned with FAIR data principles, are necessary for building harmonized datasets to train predictive models.

As composites expand into automotive, hydrogen storage, wind energy, and consumer products, the cost of undetected defect interactions increases. A certification regime based on single-defect thresholds may overlook synergistic defect combinations that degrade performance. A probabilistic framework could enable lighter, more efficient designs by allowing precise reliability quantification. By understanding defects as perturbation fields and integrating advanced monitoring and modeling techniques, the composites industry can enhance both safety and performance.