The single-fibre Broutman test was originally used to study fibre/matrix in
terface debonding behaviour. A competition between two failure mechanisms w
as observed during the Broutman test: fibre failure under compressive stres
ses and fibre/matrix interface debonding under transverse tensile stresses.
For systems presenting a good fibre/matrix adhesion, e.g. carbon-fibre/epo
xy composites, it was not possible to observe interface damage before compr
essive failure of the fibre. Consequently, the lest appeared to provide int
eresting compressive fragmentation data. Interfacial shear strength values
derived from this test were compared with those obtained from traditional t
ensile fragmentation tests. In addition, the results allowed some discussio
n on the compressive strength of the PAN-based carbon fibres. During testin
g, damage was detected using visual observation under polarised light and a
coustic emission (AE) monitoring. Failure mechanism separation based on AE
events was performed by using time domain parameters (amplitude and event w
idth). In particular, fast fourier transform frequency spectra of the AE wa
veforms were used as fingerprints allowing the discrimination of fibre fail
ure, matrix cracking, and parasitic 'noise'. In addition, fibre bundle tran
sverse tensile tests were performed with the PAN-based carbon fibre/epoxy s
ystems. The transverse strength, characteristic of the interface transverse
properties, showed the same trends as the compression fragmentation result
s. The interface adhesion quality assessed through scanning electron micros
cope fractography could be qualitatively related to the transverse strength
data. (C) 1999 Elsevier Science Ltd. All rights reserved.