Page 243 - Fiber Fracture

STRENGTH AND FRACTURE OF METALLIC FILAMENTS 227

Fig. 42. Persistent slip band in a fcc metal.

direction) in a TEM micrograph, the band forms a ladder that crosses the less regularly
spaced spots that result from the veins. The number of PSBs that form in a crystal
depend on the strain amplitude. Persistent glide bands are regions of easy glide and
sustain deformations up to the upper plateau strain (7.5 x In contrast to this,
the vein-covered matrix is comparatively hard. Therefore, when the strain amplitude is
increased, the number of PSB also increases, in order to accommodate the strain. The
situation as described above develops in the first lo3 to lo4 cycles in single crystals
and under certain conditions also in polycrystals. Multiple slip, however, appears to
perturb or even prevent the establishment of this well ordered dislocation structure.
Other structures, such as maze structures or cellular structures, form instead. A cellular
structure also forms when a strain amplitude is chosen beyond the upper plateau strain
or when PSBs are cycled for a prolonged duration. Nevertheless, the latter case is
difficult to achieve, since PSBs, once developed, rapidly form extrusions and intrusions
which then become sites of transgranular crack nucleation. These cracks then rapidly
propagate in the strain-hardened crystal and produce a fracture of the brittle type.
Inspection of the wire and fracture surfaces of the thick Cu wire (Fig. 35 right,
Fig. 36 left) now clearly indicate what happened during fatigue. The wire surface of
the fatigued sample is densely packed with extrusions that stem from PSBs; in addition,
many open micro-cracks are visible. Since this wire has a bamboo structure (i.e. a chain
of single crystals), there are always many grains with orientations where primary slip
on a single glide system will occur. Moreover, the formation of PSBs is known to be
relatively insensitive to the stress direction; therefore, after fatiguing no obvious cause
should prevent the formation of PSBs. The final fracture occurs by crack propagation as
is characteristic for strongly strain-hardened metals.
This strongly contrasts with the findings for the thin wire of the same type and subject
to exactly the same fatigue test. Its fatigue life is much longer, the final rupture is of
the ductile type and the wire surface remains smooth. This ductile fracture suggests that
strain hardening did not take place in the usual manner and the smooth surface indicates
that PSBs and extrusions did not develop in this wire. Two obvious explanations are that