Correlating fibre break development with fibre misalignment and resin-rich pockets using in situ holotomography

• In situ 150 nm synchrotron holotomography monitors fibre breaks under tension. • Misplaced/misaligned fibres and resin-rich pockets influence break development. • Interactions promote break clustering and microscale damage near breaks. • Clusters reveal severe matrix nonlinearity via matrix microcracks and debonds. • Local microstructural variations act as precursors to fibre break development. Understanding the mechanical influence of microstructural variations on fibre break development is crucial for reducing uncertainties in predicting the longitudinal tensile failure of unidirectional composites. In this study, the interaction between local microstructural variations and fibre breaks is monitored using in situ X-ray holotomography at 150 nm voxel size. Three distinctive microstructures are identified to drive the initiation and clustering of breaks. First, misplaced fibres within the 0°/0° interply region, exhibiting significant misorientation and intersecting multiple aligned fibres, progressively trigger multiple single breaks and a co-planar cluster at their intersections. Second, resin-rich pockets within the 0°/0° interply region influence break clustering, progressively forming a non-coplanar cluster of five breaks, accompanied by non-uniform matrix microcracks and short interfacial debonds surrounding the clusters. Third, large resin-rich pockets, locally formed in regions with misoriented fibre groups, play a critical role in driving pronounced break clustering. Their interactions exhibit severe matrix nonlinearity, as evidenced by matrix microcracks, short interfacial debonds, and damage features, including microvoids or the onset of matrix microcracks. The identified fibre break patterns, clustering behaviour, and damage associated with three microstructural cases provide new insights into how these microstructures serve as precursors to fibre break development, highlighting the importance of accounting for them in predicting longitudinal tensile failure to improve reliability.