Characterising fibre-matrix interface strength in composite materials using in-situ microscopy

Following the initiation and growth of interface damage under normal, shear and mixed loading

The bond between reinforcing fibres and the surrounding polymer matrix determines how effectively a composite transfers load, but also how damage develops at the microscale. We developed test methods that combine mechanical loading with in-situ microscopy to observe where this bond begins to fail and how the resulting cracks grow. Studying both short and continuous fibres under normal, shear and mixed loading provides a basis for comparing fibre surface treatments and characterising interface behaviour for composite modelling.

Interface damage is relevant to composite applications ranging from short-fibre components to continuous-fibre structures such as wind turbine blades. Depending on the fibre orientation and loading direction, the interface can experience opening forces, shear forces or a combination of both. Understanding its response therefore requires tests that examine these different conditions and distinguish the onset of damage from its subsequent growth. Furthermore, it allows us to determine interfacial strength, under different loadings, which could subsequently be used in micromechanical models of composites.

Figure. Damage and debonding initiation and propagation around a fiber tip in a glass fiber epoxy composite.

Observing debonding during loading

Our experiments use small specimens containing individual glass or carbon fibres embedded in a transparent epoxy matrix. A miniature tensile stage allows us to load these specimens under an optical microscope and follow the separation of fibre and matrix, known as debonding. Linking the images to the applied load reveals where cracks initiate, how far they extend and whether they grow gradually or spread rapidly.

Different specimen designs provide access to different loading conditions. Short fibres aligned with the loading direction allow us to follow debonding from the fibre tips under predominantly shear loading. Inclined and transverse short fibres reveal how combined opening and sliding affect the damage process. For continuous fibres loaded transversely, a modified cross-shaped specimen concentrates loading in a thin central region, allowing interface damage to be studied while limiting premature specimen failure that occurs in regular test methods.

Figure. Example of debonding events occurring when the interface is loaded under normal conditions in a glass fibre epoxy composite specimen.

The microscopy also provides information beyond crack length. Polarised light reveals stress concentrations and deformation in the surrounding matrix, helping us interpret how interface damage interacts with the material around it.

Distinguishing damage initiation from crack growth

The testing framework allows us to distinguish the initiation of new debonds from the propagation of existing cracks. This distinction is important when comparing fibre surface treatments, or sizings: a treatment that improves resistance to crack growth does not necessarily delay the first local damage. Our studies for example illustrated how a sizing with better overall fiber-matrix adhesion can still show debonding initiation at similar loads to a less compatible treatment.

Further information

Contact

Prof. dr. ir. Lode Daelemans (Lode.Daelemans@UGent.be)