4.3. Fiber length: short vs continuous
Although little data has been published on long fiber–ZrB2 composites,
the present experiments do not support the superiority of long-
fiber composite over short-fiber in terms of flexural strength or fracture
toughness values. Examples of load–displacement curves during
fracture toughness tests of unreinforced ZrB2 along with long- and
short fiber-reinforced composites are displayed in Fig. 8a. These curves
show typical features of materials having good chevron notch failure
behavior: pop-in, slow crack growth, and then catastrophic failure.
In the case of long continuous fiber composite, non-brittle failure is
clearly observed: the load-deflection curves show a non-linear slope,
achieve a maximum in applied load followed by a reduction, which
reflects stable crack growth up to final fracture of the composite. The
behavior of the bar following the initial fracture results in both total
and inelastic WOF values that are significantly improved over monolithic
materials or short fiber composites. The inelastic WOF was around 9, 17
and 78% of the total WOF for unreinforced ZrB2 and ZrB2 with short
or long fiber, respectively.
The fracture surface of a broken bar is shown in Fig. 8b where
fiber pull-out is evident. Although the higher amount of fiber, 25 vol.%,
the quality and efficiency of the coating in protecting the fiber and
promoting relevant toughening mechanisms, the fracture toughness
values measured by CNB testing were in the same range as short-fiber
composites. Some of the discrepancy between observed fracture
behavior and measured fracture toughness might be due to the testing
method. Evidence of fiber pull-out and a crack path that travels out of
the plane of the fracture surfaces proves that failure was not purely
mode I. The calculations for the CNB methods assume mode I failure.
Since this is not the case, the calculated KIc values are questionable.
However, the resulting composite had superior WOF values, indicating
that non-brittle failure was achieved.