JOURNAL OF MATERIALS SCIENCE LETTERS 12 (1993) 218-219 Fibre-matrix interface transverse tensile debonding P. MARSHALL, N. STONE British Aerospace plc, Sowerby Research Centre, PO Box 5, Filton, Bristol BS12 7QW, UK The strength of the bond between the reinforcing fibre and matrix in composite materials is recognized to be of great importance to the bulk properties of the laminated composite. Many techniques have been devised to assess the properties of the fibre- matrix interface and some have used laminated materials, either testing to destruction in three-point bending [1] or by non-destructive monitoring of the viscoelastic behaviour of the laminate [2]. However, these techniques do not measure prop- erties exclusive to the fibre-matrix interface. There are single-fibre tests which unambiguously measure the fibre-matrix interface bond, for example the push-out test [3], fibre fragmentation [4] and pull- out test [5]. One limitation of these methods is that the fibre is loaded axially so that only the interface shear properties can be evaluated. Although the shear properties of the interface are important, it is also desirable to know the interface tensile strength. The geometrical constraints to developing a trans- verse tensile mechanical test applicable to small- diameter reinforcing fibres are clear. Some success has been achieved with a compression specimen consisting of an hourglass-shaped resin sample with a centrally located fibre. Under compression the Poisson ratio effects concentrate a normal tensile force perpendicular to the fibre axis. In reflected light interface debonding can be detected and the interface tensile strength calculated [4]. Unfortun- ately, the technique is not easily applied to carbon fibres. We designed and tested a method for measuring the transverse tensile fibre-matrix bond strength of carbon fibres. The method exploits the conductive properties of the fibres and electromagnetic forces that can be induced when electrical current is applied. Polyacrylonitrile (PAN)-based intermediate- modulus carbon fibres of diameter 5/xm were mounted into card "window frames" with alumi- nium tape and carbon-filled adhesive providing electrical contact to the fibres (Fig. 1). Two samples of this type were placed together with a 70/zm-thick film of the thermoplastic polyetherimide (PEI) between them. A potential of 40 V was applied in parallel across the two fibres for approximately 10 s. The current flowing in the same direction along each fibre induced an attractive electromagnetic force between the fibres, forcing them towards each other and into contact with the PEI film (Fig. 2a). Resis- tive heating of the fibres caused the PEI surface to melt. Upon removing the current the PEI cooled rapidly, bonding the fibre into place. The sample 218 ~ Aluminium tape [\\\, ~x{~--~ Car bon fibre ~ Hole in card ..... Fold Fibre in card window Section through two card windows Two fibres separated by PEI film Figure 1 Sample preparation. consisted at this stage of two parallel fibres embedded in the PEI film (Fig. 2b). The procedure was conducted under an optical microscope, but the situation is illustrated better in the electron micro- graph shown in Fig. 3, where a single side of the sample can be seen in detail. The tensile force needed to break the interface bond was provided by passing a short current pulse in opposite directions along the two fibres, which generated an instantan- eous repulsive electromagnetic force (Fig. 2c). A field-effect transducer ensured a current pulse decay of <20/xs, which minimized the resistive heating effects. The power supply output was increased in steps, providing incremental increases in the electro- magnetic force until the fibres became debonded. Fig. 4 shows an electron micrograph of a debonded fibre; note the thin layer of PEI adhering to the fibre surface. The force acting between the fibres at debonding, F, can be calculated from F = IxliI21/27ra 0261-8028 © 1993 Chapman & Hall